Decision 22570-D01-2018 2018 Generic Cost of Capital · 2018-08-03 · Decision 22570-D01-2018...

Decision 22570-D01-2018

2018 Generic Cost of Capital August 2, 2018

Alberta Utilities Commission

Decision 22570-D01-2018

2018 Generic Cost of Capital

Proceeding 22570

August 2, 2018

Published by the:


Eau Claire Tower, 1400, 600 Third Avenue S.W.

Calgary, Alberta

T2P 0G5

Telephone: 403-592-8845

Fax: 403-592-4406

Website: www.auc.ab.ca

Decision 22570-D01-2018 (August 2, 2018) • i

Contents

1 Introduction ................................................................................................................. 1

2 Procedural summary .................................................................................................. 2

3 Overview of the Commission’s approach to setting an approved ROE and

approved deemed equity ratios .................................................................................. 4

4 Fair return standard ................................................................................................... 6

5 Income taxes .............................................................................................................. 10 5.1 Standardization of income tax methodology ........................................................... 11

5.1.1 Relevant factors in assessing the suitability of an income tax method ........ 13

5.2 Claiming maximum allowable income tax deductions when forecasting income tax

expense ..................................................................................................................... 19

5.3 Reporting of future income tax liabilities ................................................................ 19 5.4 Income tax deferral accounts ................................................................................... 20

5.5 PBR implications ..................................................................................................... 26

6 Relevant changes in global economic and Canadian capital market conditions

since the 2016 GCOC decision ................................................................................. 27 6.1 Macroeconomic conditions ...................................................................................... 28 6.2 Inflation .................................................................................................................... 30

6.3 Interest rate environment ......................................................................................... 31 6.4 Credit spreads........................................................................................................... 33

6.5 Market volatility....................................................................................................... 37

6.6 Overall conclusions of the witnesses ....................................................................... 40

6.7 Commission findings ............................................................................................... 41

7 Municipally owned utilities ...................................................................................... 46 7.1 Ownership structure and stand-alone principle ........................................................ 46 7.2 ACFA funding, credit metrics and stand-alone principle ........................................ 50 7.3 Use of equity funding riders .................................................................................... 53

8 Return on equity ....................................................................................................... 54 8.1 Use of proxy group companies ................................................................................ 55 8.2 The capital asset pricing model................................................................................ 59

8.2.1 Risk-free rate................................................................................................ 60

8.2.2 Market equity risk premium ........................................................................ 65 8.2.3 Beta .............................................................................................................. 70 8.2.4 The resulting CAPM estimate...................................................................... 75 8.2.5 The ECAPM ................................................................................................ 76

8.3 Other risk premium models ..................................................................................... 80 8.4 Discounted cash flow model .................................................................................... 85 8.5 Stock market return expectations of finance professionals...................................... 94 8.6 Flotation allowance .................................................................................................. 98 8.7 Other considerations in establishing a fair approved return on equity .................... 98 8.8 Conclusions on ROE .............................................................................................. 100 8.9 Returning to a formula-based approach to establishing ROE ................................ 104

ii • Decision 22570-D01-2018 (August 2, 2018)

9 Capital structure matters ....................................................................................... 105 9.1 Overview ................................................................................................................ 105

9.2 Deemed equity ratios requested ............................................................................. 106 9.3.1 Overall assessment of business risk ........................................................... 111 9.3.2 Changes in business risk since the 2016 GCOC decision ......................... 111

9.3.2.1 2018-2022 PBR term ....................................................................... 111

9.3.2.2 The Commission’s UAD decision and the related issue of asset

utilization ......................................................................................... 115 9.3.2.3 Increase in customer contributions .................................................. 119 9.3.2.4 Regulatory lag .................................................................................. 121 9.3.2.5 Clean energy initiatives ................................................................... 122

9.3.3 Business risk comparisons between the affected utilities and other

jurisdictions ................................................................................................ 123 9.3.4 Comparability of deemed equity ratios ...................................................... 131

9.4 Industry financing practices ................................................................................... 134 9.5 Factors raised by FortisAlberta .............................................................................. 135 9.6 Maintaining credit ratings in the A-range .............................................................. 136

9.7 Credit ratings and credit metric analysis ................................................................ 140 9.7.1 Financial ratios, capital structure and actual credit ratings ........................ 140 9.7.2 Equity ratios associated with credit metrics .............................................. 151

9.8 Business risk utility sector analysis ....................................................................... 159 9.9 Equity ratio adjustments for income-tax-exempt or non-taxable utilities ............. 161

9.10 Equity ratio adjustments for ENMAX ................................................................... 163 9.11 Determination of Commission-approved deemed equity ratios ............................ 164

10 Determination of Commission-approved deemed equity ratio for AltaGas

Utilities Inc. .............................................................................................................. 165

11 Other areas included in scope of the proceeding ................................................. 171 11.1 Maintaining actual equity ratio in line with deemed equity ratio .......................... 171 11.2 Reporting of information in Rule 005 .................................................................... 172

11.3 Procedural issues .................................................................................................... 174 11.3.1 Use of aids to cross-examination ............................................................... 174

11.3.2 UCA concerns regarding process .............................................................. 175

12 Implementation of GCOC decision findings ........................................................ 176

13 Order ........................................................................................................................ 178

Appendix 1 – Proceeding participants .................................................................................... 181

Appendix 2 – Oral hearing – registered appearances ........................................................... 183

Appendix 3 – Summary of Commission directions ................................................................ 185

Appendix 4 – Abbreviations ..................................................................................................... 187

Decision 22570-D01-2018 (August 2, 2018) • iii

List of tables

Table 1. Approved final ROE for 2018, 2019 and 2020, and approved final deemed equity

ratios for 2018, 2019 and 2020 ................................................................................... 2

Table 2. Forward-looking expected MERPs as reported by Mr. Coyne ............................ 67

Table 3. Forward-looking expected MERPs as reported by Mr. Hevert ........................... 68

Table 4. CAPM inputs and resulting ROE estimates ........................................................... 75

Table 5. ROE results determined using various DCF models, including flotation

allowance .................................................................................................................... 91

Table 6. ROE recommendations presented ......................................................................... 100

Table 7. Currently approved deemed equity ratios and the deemed equity ratios

recommended for 2018, 2019 and 2020 ................................................................. 109

Table 8. Parameters for calculating credit metrics ............................................................ 152

Table 9. Embedded average debt rates, depreciation rates and CWIP percentages by

utility ........................................................................................................................ 153

Table 10. Additional analysis of information included in Table 9 ...................................... 154

Table 11. Credit metrics compared to equity ratios – Commission calculations –

distribution utilities – income tax rate of 27 per cent .......................................... 156


distribution utilities – income tax rate of zero...................................................... 157


transmission utilities – income tax rate of 27 per cent ........................................ 157


transmission utilities – income tax rate of zero .................................................... 158

Table 15. Commission guidelines for equity ratios to achieve a credit rating in the A-range

................................................................................................................................... 159

List of figures

Figure 1 Yield curves for GOC and U.S. government bonds as of March 26, 2018 .......... 31

Figure 2 Canadian government bond yields, 10-year vs. 30-year ........................................ 32

Figure 3 30-year Canadian A-rated utility bond yields, 30-year GOC bond yields........... 35

iv • Decision 22570-D01-2018 (August 2, 2018)

Figure 4 Credit spread between 30-year Canadian A-rated utility bond yields and 30-year

GOC bond yields ....................................................................................................... 35

Figure 5 30-year credit spreads for Alberta utilities ............................................................. 36

Figure 6 Canadian and U.S. stock market volatility indexes ............................................... 39

Figure 7 10-year, 30-year GOC bonds and the 30-year utility bond yields ........................ 44

Decision 22570-D01-2018 (August 2, 2018) • 1


Calgary, Alberta

Decision 22570-D01-2018

2018 Generic Cost of Capital Proceeding 22570

1 Introduction

1. This decision sets out the approved return on equity (ROE) for the years 2018, 2019 and

2020 on a final basis. The approved ROE applies uniformly to the utilities listed below:

AltaGas Utilities Inc. (AltaGas)1

AltaLink Management Ltd. (AltaLink)2

ATCO Electric Ltd. (ATCO Electric)3

ATCO Gas & Pipelines Ltd.4 5

ENMAX Power Corporation (ENMAX)6

EPCOR Distribution & Transmission Inc. (EPCOR)7

FortisAlberta Inc. (FortisAlberta)8

City of Lethbridge (Lethbridge)9

City of Red Deer (Red Deer)10

TransAlta Corporation (TransAlta)11

(collectively, the affected utilities)

2. This decision also sets out the approved deemed equity ratios (also referred to as capital

structure) for the affected utilities for 2018, 2019 and 2020 on a final basis.

3. Additionally, this decision considers the two commonly used income tax methodologies,

flow-through and future income tax (FIT), and whether the Alberta Utilities Commission will

direct the adoption of one standard methodology.

4. The approved final ROE and final deemed equity ratios for 2018, 2019 and 2020 for all

of the affected utilities are set out in Table 1 below.

1 Natural gas distribution. 2 Electricity transmission. 3 Electricity transmission and distribution. Unless otherwise indicated, a reference to ATCO Electric includes

both the transmission and distribution operations of this utility. 4 ATCO Gas refers to the utility’s natural gas distribution operations. ATCO Pipelines refers to the utility’s

natural gas transmission operations. 5 Collectively, ATCO Electric, ATCO Gas and ATCO Pipelines are referred to as the ATCO Utilities. 6 Electricity transmission and distribution. Unless otherwise indicated, a reference to ENMAX refers to both the

transmission and distribution operations of this utility. 7 Electricity transmission and distribution. Unless otherwise indicated, a reference to EPCOR refers to both the

transmission and distribution operations of this utility. 8 Electricity distribution. 9 Electricity transmission. 10 Electricity transmission. 11 Electricity transmission assets.


2 • Decision 22570-D01-2018 (August 2, 2018)

Table 1. Approved final ROE for 2018, 2019 and 2020, and approved final deemed equity ratios for 2018, 2019 and 2020

2018 approved 2019 approved 2020 approved

(%)

ROE 8.5 8.5 8.5

Deemed equity ratios

Electricity and natural gas transmission

AltaLink 37 37 37

ATCO Electric 37 37 37

ATCO Pipelines 37 37 37

ENMAX 37 37 37

EPCOR 37 37 37

Lethbridge 37 37 37

Red Deer 37 37 37

TransAlta 37 37 37

Electricity and natural gas distribution

AltaGas 39 39 39

ATCO Electric 37 37 37

ATCO Gas 37 37 37

ENMAX 37 37 37

EPCOR 37 37 37

FortisAlberta 37 37 37

5. The approved ROE and deemed equity ratios from this decision do not apply to EPCOR

Energy Alberta GP Inc., ENMAX Energy Corporation and Direct Energy Regulated Services

because these utilities are regulated pursuant to the Electric Utilities Act, Regulated Rate Option

Regulation12 and the Gas Utilities Act Default Gas Supply Regulation,13 respectively.

6. The ROE and deemed equity ratios for the various investor-owned water utilities under

the Commission’s jurisdiction were not determined in this proceeding. However, the

determinations in this proceeding may be considered in other proceedings, should issues

respecting ROE and deemed equity ratios arise for these utilities.

2 Procedural summary

7. On October 7, 2016, the Commission issued Decision 20622-D01-201614 (2016 Generic

Cost of Capital (GCOC) decision), which set an approved ROE and deemed equity ratios for

2016 and 2017. With respect to 2018, the Commission stated:

12 Alberta Regulation 262/2005. 13 Alberta Regulation 184/2003. 14 Decision 20622-D01-2016: 2016 Generic Cost of Capital, Proceeding 20622, October 7, 2016.



339. The allowed ROE for 2017 of 8.50 per cent awarded in this decision will remain

in place on an interim basis for 2018 and for subsequent years until changed by the

Commission.15

…

623. The approved deemed equity ratios awarded in this decision will remain in place

on an interim basis for 2018 and for subsequent years until changed by the Commission.16

8. On April 20, 2017, in Proceeding 20687: Commission-Initiated Generic Proceeding to

Address the Income Tax Methodologies Used in Revenue Requirement Calculations for

Regulated Utilities in Alberta (Proceeding 20687), the Commission issued correspondence

indicating that it was not prepared to prescribe a single income tax methodology without first

examining the implications of changing tax methodologies on other components that may affect

the setting of rates by the Commission, such as cost of capital.17 The Commission stated that the

best forum to consider these matters was the 2018 GCOC proceeding. Accordingly, the

Commission stated that the record of Proceeding 20687 would form part of the record of the

2018 GCOC proceeding.

9. Also on April 20, 2017, the Commission issued a letter initiating the 2018 GCOC

proceeding, Proceeding 22570.18 That letter proposed the scope of issues to be considered in

Proceeding 22570 as well as process timelines and provided interested parties with an

opportunity to comment.

10. On July 5, 2017, the Commission ruled that it would establish approved ROEs and

deemed equity ratios for 2018, 2019 and 2020 in Proceeding 22570. The Commission also

addressed the scope of the proceeding and the minimum filing requirements for the utilities.

The Commission identified that the scope of Proceeding 22570 would include:

Whether changes in the approved ROE and deemed equity ratios established in the

2016 GCOC decision are warranted.

How the Commission should consider the traditional approaches and models used in

previous GCOC proceedings for determining an approved ROE and equity ratios.

The short-term and long-term effects of employing the two commonly used income

tax methodologies, flow-through and FIT, on areas such as cost of capital and overall

revenue requirement, and how the Commission should consider factors such as

differences in the sum of the present discounted value of the revenue requirement and

impacts on funds from operations (FFO)/debt in deciding which method should be

applied to utilities.19

The issues surrounding income tax methods or treatments, income tax deferral

accounts, and performance-based regulation (PBR) implications, as set out by the

Commission in its issues list in Proceeding 20687.20

Relevant issues regarding long-term debt.21

15 Decision 20622-D01-2016, paragraph 339. 16 Decision 20622-D01-2016, paragraph 623. 17 Exhibit 22570-X0077, paragraph 11. 18 Exhibit 22570-X0078. 19 Exhibit 22570-X0114, paragraph 28. 20 Exhibit 22570-X0114, paragraph 29.



Matters with respect to municipally owned utilities, specifically how their ownership

structure and the relationship between utility ratepayers and municipal taxpayers may

affect ROE and deemed equity ratios for these utilities.22

11. Each of the affected utilities, except Lethbridge, Red Deer and TransAlta, actively

participated in this proceeding. AltaGas and the ATCO Utilities co-sponsored the evidence of

Dr. Bente Villadsen, Dr. Paul Carpenter and Mr. Robert Buttke. AltaLink, EPCOR and

FortisAlberta co-sponsored the evidence of Mr. Robert Hevert. ENMAX sponsored the evidence

of Mr. James Coyne. Additionally, each of AltaGas, AltaLink, ENMAX, EPCOR, FortisAlberta

and the ATCO Utilities (collectively, the utilities) filed company-specific evidence, including the

minimum filing requirements directed by the Commission.23

12. The City of Calgary (Calgary), the Consumers’ Coalition of Alberta (CCA) and the

Office of the Utilities Consumer Advocate (UCA) (collectively, the interveners) actively

participated in the proceeding. Calgary sponsored the evidence of Mr. Hugh Johnson; the CCA

submitted the evidence of Mr. Jan Thygesen and Mr. Dustin Madsen; and the UCA sponsored

the evidence of Dr. Sean Cleary and Mr. Russ Bell.

13. In addition to the filing of evidence, the Commission’s process included information

requests (IRs) and responses on the utilities’ evidence, and evidence sponsored by the utilities;

IRs and responses on evidence sponsored by the interveners; rebuttal evidence filed by the

utilities; a two-week oral hearing; and a further process to permit IRs and responses to follow up

on outstanding answers to undertakings. The Commission also established a process for

simultaneous written argument and reply argument.

14. The Commission considers that the record of this proceeding closed with the filing of

reply arguments on May 8, 2018.

15. In reaching the determinations set out in this decision, the Commission has considered all

relevant materials comprising the record of this proceeding, including the evidence and argument

provided by each party, and the evidence and submissions from Proceeding 20687. Accordingly,

references in this decision to specific parts of the record are intended to assist the reader in

understanding the Commission’s reasoning relating to a particular matter and should not be taken

as an indication that the Commission did not consider all relevant portions of the record with

respect to that matter.

3 Overview of the Commission’s approach to setting an approved ROE and

approved deemed equity ratios

16. In satisfying the fair return standard, the Commission is required to determine a fair ROE

for the affected utilities. In Decision 2009-21624 (2009 GCOC decision), Decision 2011-47425

21 Exhibit 22570-X0114, paragraph 34. 22 Exhibit 22570-X0114, paragraph 36. 23 A complete list of registered participants is produced in Appendix 1 to this decision. 24 Decision 2009-216: 2009 Generic Cost of Capital, Proceeding 85, Application 1578571-1, November 12, 2009,

paragraphs 77-78. 25 Decision 2011-474: 2011 Generic Cost of Capital Proceeding, Proceeding 833, Application 1606549-1,

December 8, 2011, paragraph 2.



(2011 GCOC decision), Decision 2191-D01-201526 (2013 GCOC decision) and the 2016 GCOC

decision,27 the Commission established an ROE that uniformly applied to all of the affected

utilities and accounted for particular business risks faced by the affected utilities by

incorporating any required adjustments into their respective approved deemed equity ratios,

either collectively or on an individual basis. The Commission adopted the same approach in this

decision.

17. For the purposes of this decision, the Commission’s point of departure is the approved

ROE and deemed equity ratios established in the 2016 GCOC decision. From this starting point,

the Commission has evaluated the evidence and argument in this proceeding to determine

whether changes in the approved ROE and deemed equity ratios from the 2016 GCOC decision

are warranted.

18. In determining a fair ROE, the Commission begins, in Section 4, with a discussion of the

fair return standard. This is followed by a discussion of income tax in Section 5.

19. In Section 6, the Commission evaluates changes in the global economic and Canadian

capital market conditions since the conclusion of the 2016 GCOC proceeding. This review is

a factor informing the Commission’s determination of both a fair approved ROE and deemed

equity ratios, as discussed in sections 8 and 9.

20. In Section 7, the Commission considers issues related to the municipally owned utilities,

including the availability of the Alberta Capital Financing Authority (ACFA) financing and

equity funding riders.

21. In Section 8, the Commission establishes the approved ROE for 2018, 2019 and 2020 on

a final basis, after consideration of all the relevant factors, including changes in global economic

and Canadian capital market conditions, financial models and the effect of potential regulatory

risk factors identified by parties.

22. In Section 9, the Commission establishes the approved deemed equity ratios for 2018,

2019 and 2020, for all of the affected utilities other than AltaGas, after consideration of all the

relevant factors, including credit metric analysis, business risk analysis, generic business risks,

utility sector business risk analysis and any company specific adjustments. In Section 10, the

Commission establishes the approved deemed equity ratio for AltaGas.

23. In Section 11, the Commission addresses other issues raised during the proceeding that

are not specifically addressed in other sections.

24. In Section 12 of the decision, the Commission sets out how the approved ROE and

deemed equity ratios are to be implemented by the affected utilities.

26 Decision 2191-D01-2015: 2013 Generic Cost of Capital, Proceeding 2191, Application 1608918-1, March 23,

2015, paragraph 416. 27 Decision 20622-D01-2016, paragraph 340.



4 Fair return standard

25. All parties agreed that the fair return standard requires consideration of three factors,

specifically ‘”comparable investments,” “capital attraction” and “financial integrity.” The CCA

suggested that, in addition to these three factors, the fair return should consider ratepayer

impacts, while many of the utilities argued that the fair return should only be considered from the

perspective of the utility equity investor. Another point of disagreement between the utilities and

interveners related to the weight to be placed on comparable investments.

26. The CCA submitted that the Commission should have regard to the impact of any

approved ROE and equity thickness on both customers and utilities, and that the fair return

should be no more than is absolutely required to maintain safe, reliable and economic service for

the foreseeable future.28 In its reply argument, the CCA argued that the fair return standard does

not override the requirement that rates be just and reasonable.29

27. EPCOR argued that the fair return factors, assessed from the perspective of an investor,

address both investor and customer interests as customers benefit from being served by a

functional utility that is able to maintain its financial integrity and attract capital.30 In its reply

argument, EPCOR discussed how an ROE of 25 per cent would doubtlessly satisfy the financial

integrity and capital attraction factors, but would be too high to satisfy the comparable

investment component, which operates to ensure that the utility and customers pay no more for

equity than what the market requires.31 EPCOR agreed that the just and reasonable standard is

a fundamental legal requirement that applies to utility rates generally and to the fair return in

particular.32

28. The ATCO Utilities and AltaGas suggested that the CCA’s recommendation was an

attempt to fabricate a new test for the fair return standard, and referred to several statements from

TransCanada Pipelines Ltd v National Energy Board, including:

… While I agree with the appellant that the impact on customers or consumers cannot be

a factor in the determination of the cost of equity capital, any resulting increase in tolls

may be a relevant factor for the Board to consider in determining the way in which a

utility should recover its costs. It may be that an increase is so significant that it would

lead to “rate shock” if implemented all at once and therefore should be phased in over

time. It is quite proper for the Board to take such considerations into account, provided

that there is, over a reasonable period of time, no economic loss to the utility in the

process.33

29. ENMAX argued that the revenue requirement impact on customers is not a relevant

consideration in determining a fair return.34 Similarly, the ATCO Utilities and AltaGas, in their

joint argument, stated that the impact on customer rates is irrelevant when determining the

28 Exhibit 22570-X0888, paragraph 91. 29 Exhibit 22570-X0920, paragraph 223. 30 Exhibit 22570-X0893, paragraph 65. 31 Exhibit 22570-X0915, paragraph 41. 32 Exhibit 22570-X0915, paragraph 34. 33 Exhibit 22570-X0918, paragraph 27. 34 Exhibit 22570-X0896, paragraph 19.



required rate of ROE and that other regulatory mechanisms are available to mitigate impacts on

customers.35

30. With respect to comparability, in his oral testimony, Mr. Coyne reflected that “I believe

that the Commission in 2016 took a leg out from under the stool, or at least shortened it when it

put greater reliance on just the credit rating.”36 In his rebuttal evidence, Mr. Coyne provided a

figure showing the approved equity returns of Canadian gas and electric distribution utilities that

have rates set through a litigated proceeding.37 In oral testimony, Mr. Coyne admitted that this

figure does not adjust for risk.38 Mr. Coyne described this figure as an objective measure of what

comparability looks like and that it is one valid way to consider all three fair return factors.39

ENMAX argued that if the conclusion is that Alberta utilities are average risk compared to other

Canadian utilities, then they should be in the middle of the figure to meet the comparable

investments factor.40

31. In her evidence, Dr. Villadsen provided a summary of approved ROE and capital

structures for regulated Canadian and United States (U.S.) utilities, which she submitted were

relevant because investors compare returns across jurisdictions.41

32. In argument, EPCOR focused on the comparability or comparable investment

component:

57. As developed by subsequent Canadian authorities, this aspect of the “fair return”

standard has found its most complete expression in the “comparability” or “comparable

investment” component of the test. This component has variously been expressed as

requiring “(t)hat the investor should be able to obtain a return from his investment such

as might alternatively be obtained from other investments of comparable risk and

uncertainty,” or that a fair return “be comparable to the return available from the

application of the invested capital to other enterprises of like risk”.42

33. EPCOR argued that a return that does not satisfy the comparability standard would not

allow a utility to raise new capital or engage in refinancing.43

34. Mr. Hevert stated that the required return is a function of the risk and return

characteristics of the investment, and not the source of the funds.44 EPCOR noted that Dr. Cleary

confirmed this principle.45 Mr. Hevert submitted that any notion of a company having a different

value depending on how its investors fund their equity investment violates the widely

35 Exhibit 22570-X0900, paragraph 26. 36 Transcript, Volume 5, page 1002. 37 Exhibit 22570-X0775, PDF page 40. 38 Transcript, Volume 5, page 1005. 39 Transcript, Volume 5, page 1005. 40 Exhibit 22570-X0896, paragraph 47. 41 Exhibit 22570-X0193.01, PDF pages 77-78. 42 Exhibit 22570-X0893, paragraph 57. 43 Exhibit 22570-X0893, paragraph 60. 44 Exhibit 22570-X0153.01, PDF pages 16, 125-126. 45 Transcript, Volume 10, page 2161.



acknowledged economic “law of one price.” He added this principal states that in an efficient

market, identical assets have the same value.46

Commission findings

35. In each of the Public Utilities Act, the Gas Utilities Act and the Electric Utilities Act, the

fair return is referenced as a component of just and reasonable rates. The Public Utilities Act and

Gas Utilities Act require the Commission, in fixing just and reasonable rates, to determine a rate

base upon which it shall fix a fair return.47 The Electric Utilities Act requires the Commission to

ensure that a tariff is just and reasonable and provides the owner of an electric utility with a

reasonable opportunity to recover a fair return on the equity of shareholders of the electric

utility.48

36. The interplay between just and reasonable rates and a fair return was described in

Northwestern Utilities Ltd. v Edmonton (City) as follows:

The duty of the Board was to fix fair and reasonable rates; rates which, under the

circumstances, would be fair to the consumer on the one hand, and which, on the other

hand, would secure to the company a fair return for the capital invested.49

37. As reflected in the preceding quotation, determining just and reasonable rates balances

the interests of both the utility and its customers. The Commission exercises its judgment in

determining a total return for each utility to establish rates that provide the utility a reasonable

opportunity to earn a fair return on invested capital while ensuring that rates are just and

reasonable so that customers are not paying more than is required to maintain safe, reliable and

economic service.

38. The approach to determining a fair return on the equity component of invested capital in

a regulated utility has ordinarily been referred to as the fair return standard.50 The Commission

has addressed the fair return standard in previous GCOC decisions, with Decision 2009-21651

providing a thorough discussion of the underlying statutory framework and relevant case law.

As discussed in Decision 2009-216, the Commission and its predecessors have accepted and

considered the following three factors when setting a fair return: “comparable investments,”

“capital attraction” and “financial integrity.” The Commission considers these factors to be well

established and continues to be satisfied that the fair return standard is met when the return

satisfies these three factors, while also understanding that this is a component of just and

reasonable rates.

39. The Commission also does not consider that simply matching the ROE and deemed

equity ratios awarded by other regulators satisfies the fair return standard, nor does it establish

just and reasonable rates. The Commission remains of the view that seeking to match approved

returns in other jurisdictions provides an outcome that is inherently circular. The objective of the

46 Exhibit 22570-X0153.01, PDF page 126. 47 Public Utilities Act, RSA 2000, c P-4, s 90(1); Gas Utilities Act, RSA 2000, c G-5, s 37(1). 48 Electric Utilities Act, SA 2003, c E-5.1, s 121(2)(a), 122(1)(a)(iv). 49 Northwestern Utilities Ltd v Edmonton (City), 1929 CarswellAlta 114, paragraph 18, [1929] 2 DLR 4, [1929]

SCR 186. 50 Decision 2009-216, paragraph 87. 51 Decision 2009-216, Section 2.



GCOC is to consider the market expectation for the utilities raising capital and providing utility

service in Alberta, not simply mimicking the returns awarded by other regulators.

40. In addition, there is insufficient evidence to conclude that utilities in other Canadian and

U.S. jurisdictions are comparable and face the same risks as the affected utilities. The

determination of a “comparable” return requires the Commission to apply its judgment in

assessing the specific cost of capital for the utilities based on the evidence presented and in the

absence of any utility under the Commission’s jurisdiction issuing equity directly to investors.

41. The Commission acknowledges Mr. Hevert’s view that “The opportunity cost concept

applies regardless of the source of the funding.” and that the nature of the investor does not

necessarily impact the expected risk-adjusted return of an investment.52 However, the

Commission operates within its legislated mandate, and does not take this principle to mean that

the owner or investor in the regulated utility must be disregarded in all contexts. Where the

source of funds is likely to result in harm to customers of a regulated utility, the Commission

may consider this.

42. For example, in circumstances where the Commission is tasked with approving the sale

of a utility to another investor, the Commission has traditionally applied a no-harm test that

considers, amongst other things, the impact of such a sale on customer rates and the impact on

the financial profile of the utility for the purposes of attracting capital.53

43. The Commission may deny a sale or other transaction under sections 101 and 102 of the

Public Utilities Act if the Commission determines that the transaction is likely to result in harm

to customers in terms of the rates paid for service or the reliability of that service.

44. In applying for approval of a multi-step transaction whereby MidAmerican (Alberta)

Canada Holdings Corporation (MC Alberta), a wholly owned subsidiary of Berkshire Hathaway

Energy Company (BHE), replaced SNC as the owner of the entities that own and operate

AltaLink, L.P.’s (ALP) transmission assets and business, MC Alberta cited the following passage

made by the Alberta Energy and Utilities Board (the Board), the Commission’s predecessor:

The Board notes that, with respect to these types of applications, any potential benefits

are generally intangible and harder to quantify. The Board notes that one persuasive

factor in any sale is that a company that wants to be in the business is replacing one that

wishes to exit the business.54

45. In its decision approving the above-noted transaction, the Commission noted:

116. In addition, MC Alberta noted that in its credit rating analysis, S&P took specific

note of the fact that utility ownership is a core business of BHE, and also suggested that

the fact that AILP [AltaLink Investments, L.P.] would be of more strategic importance to

52 Exhibit 22570-X0153.01, PDF page 16. 53 Decision 2014-326: AltaLink Investment Management Ltd. and SNC Lavalin Transmission Ltd. et al.,

Proposed Sale of AltaLink, L.P. Transmission Assets and Business to MidAmerican (Alberta) Canada Holdings

Corporation, Proceeding 3250, Applications 1610595-1, 1610596-1, 1610597-1, November 28, 2014,

paragraphs 107-108. 54 Decision 2014-326, paragraph 115.



BHE, as compared to the nonstrategic status of ALP to SNC-Lavalin could affect credit

ratings after the close of the proposed transaction.

117. MC Alberta further noted that the Commission’s predecessor had ascribed a

benefit to customers from the acquisition of a utility by an owner with “access to

extensive experience … through its affiliated companies” and argued that the fact that

BHE affiliated companies have extensive experience in the electric utility industry is a

relevant consideration for the Commission.55

46. The Commission recognized in Decision 2014-326 that the source of funds / financial

strength of the owner is a consideration in determining whether a transaction satisfies the “no

harm” test, and that any impact on credit ratings may have a corresponding impact on rates.56

The Commission found that the willingness, experience and financial strength of the proposed

owner of AltaLink was a positive factor.

5 Income taxes

47. As noted in Section 2, the scope of this proceeding includes various issues originally

identified in the Commission-initiated generic proceeding on income taxes (Proceeding 20687).

In a letter dated April 20, 2017,57 the Commission decided that the income tax issues should be

addressed as part of the 2018 GCOC proceeding. The Commission therefore closed Proceeding

20687 and informed parties that the record of Proceeding 2068758 would form part of the record

of this GCOC proceeding.

48. The scope of Proceeding 20687 included a consideration of income tax methods, income

tax deferral accounts and PBR implications. The scope was expanded in this GCOC proceeding

to include a consideration of the effects of employing flow-through and FIT on areas such as cost

of capital and overall revenue requirement.59

49. The Commission addresses the following issues regarding income tax in the sections that

follow:

Income tax methods, including whether all the utilities should adopt one standard

method.

Claiming maximum allowable income tax deductions when forecasting income tax

expense.

The CCA’s recommendation regarding reporting future income tax liabilities.

Use of deferral accounts for income tax.

PBR implications associated with income tax.

55 Decision 2014-326, paragraphs 116-117. 56 Decision 2014-326, paragraphs 122-123. 57 Exhibit 22570-X0077. 58 Exhibits 22570-X0002 to 22570-X0077 comprise the record of Proceeding 20687. 59 Exhibit 22570-X0078, paragraph 3.



50. AltaGas, AltaLink, the ATCO Utilities and FortisAlberta (the taxable utilities) are subject

to income tax, although FortisAlberta is currently in a non-tax-paying position as a result of

maximizing allowable deductions for income tax.60 As municipally owned utilities, EPCOR,

ENMAX, Lethbridge and Red Deer are exempt from paying income taxes.61

5.1 Standardization of income tax methodology

51. For all taxable utilities, the currently approved method for determining the forecast

income taxes to be included in revenue requirement is the flow-through method, with one

exception. While the provincial income taxes for ATCO Electric Transmission are determined

using the flow-through method, the federal income taxes are determined using the FIT method.

As described by the ATCO Utilities, the Commission approved the use of the FIT method for

federal income taxes for ATCO Electric Transmission to support its credit metrics at sufficient

levels to target credit ratings in the A-range.62

52. In this proceeding, parties focussed on two common income tax methods: the flow-

through method and the FIT method.

53. The flow-through method is analogous to the cash basis of accounting. When using the

flow-through method, the forecast income tax is calculated by multiplying the forecast income

tax rates (federal and provincial) by the respective federal and provincial taxable income. In

determining taxable income, non-cash expenses such as depreciation are not deductible. Instead

of depreciation, the taxing authorities permit a deduction called capital cost allowance. The

depreciation rates approved by the Commission are generally lower than the capital cost

allowance rates approved by the federal and provincial taxing authorities. Consequently, during

periods when the monetary value of the capital asset additions of a utility is quite large, the

capital cost allowance deduction is much greater than the non-deductible depreciation expense.

54. In addition, income tax deductions are available for items such as overhead costs

associated with capital assets. While these overhead costs are capitalized and recovered from

customers over the life of the capital asset for utility ratemaking purposes, the costs are fully

deductible for income tax purposes in the year they are incurred.

55. As a result of differences in depreciation and capital cost allowance and the ability to

immediately deduct certain costs for income tax purposes, taxable income and income taxes are

lower in periods when the utility has capital asset additions of significant monetary value.

56. The flow-through method permits the utility to take advantage of all available income tax

deductions. As discussed above, this helps reduce income taxes during periods of significant rate

base additions. However, the flow-through method does not include any recognition for future

periods when the capital cost allowance pools may be diminished. Any diminished capital cost

allowance pools in future periods would result in increased income taxes in those future periods,

all else being equal.

60 Exhibit 22570-X0039, paragraph 12 61 As noted in Exhibit 22570-X0002, paragraph 3: “… a municipal corporation that earns more than 90 per cent of

its income within the geographical boundaries of the municipality is exempt from paying income taxes pursuant

to the Income Tax Act …” 62 Exhibit 22570-X0900, paragraph 228.



57. The FIT method is analogous to the accrual basis of accounting and consists of two

components. The first component is the cash income taxes. The cash component, being the

amount that would have to be paid to the taxing authorities, is determined using the flow-through

method described above. The second component is the future income taxes. The future income

taxes are determined by accounting for all the differences between the non-cash expenses and the

income tax deductions. Because these differences are accounted for, the FIT method recognizes

the liability for increased income taxes in future periods, all else being equal. Any FIT balances

are also adjusted for changes in future income tax rates.63

58. All of the taxable utilities have had substantial capital asset additions over the last

number of years. Consequently, the income taxes calculated for the taxable utilities using the

flow-through method are lower than if the FIT method had been used. FortisAlberta submitted

that its continued use of the flow-through method is the most advantageous approach for both

customers and FortisAlberta.64 AltaGas,65 AltaLink66 and the ATCO Utilities67 all submitted that

the flow-through method should be used absent any special circumstances. Among the special

circumstances identified by the ATCO Utilities were credit metric support during periods of

large capital growth, a change in credit metric targets, and consistency with accounting

standards.68 AltaLink mentioned the circumstance of imminent aggregate cross-over, which

would occur when the available income tax deductions are less than the non-cash deductions.69

AltaGas indicated that the flow-through method is commonly used by other regulated utilities in

Canada.70

59. Among the interveners, Calgary and the UCA supported the use of the flow-through

method. Calgary indicated this is consistent with decisions of the National Energy Board and the

Ontario Energy Board, is beneficial to the customers of the utilities and reduces risk to the

utilities.71 The UCA supported the position of Mr. Bell that the flow-through method be used,

absent any special circumstances such as imminent cross-over or the downgrade of credit

metrics.72

60. The only party opposed to the continued use of the flow-through method was the CCA.

Based on the evidence of Mr. Madsen, the CCA submitted that the FIT method be approved as

the preferred method for accounting for income taxes.73 However, it cautioned that the

assessment of whether it is appropriate for a utility to immediately transition to FIT must be done

on a utility-by-utility basis.74

63 Exhibit 22570-X0044, PDF page 12. 64 Exhibit 22570-X0228, paragraph 8. 65 Exhibit 22570-X0127, paragraphs 16 and 21. 66 Exhibit 22570-X0141, paragraph 53. 67 Exhibit 22570-X0171, paragraph 4. 68 Exhibit 22570-X0171, paragraph 6.. 69 Exhibit 22570-X0141, paragraph 53. 70 Exhibit 22570-X0127, paragraph 9. 71 Exhibit 22570-X0903, paragraph 56. 72 Exhibit 22570-X0897.01, paragraph 359. Exhibit 22570-X0050, paragraph 13. 73 Exhibit 22570-X0888, paragraph 469. 74 Exhibit 22570-X0888, paragraph 393.



5.1.1 Relevant factors in assessing the suitability of an income tax method

61. Mr. Madsen based his recommendation for the adoption of the FIT method on four

principles: intergenerational equity, matching, cost causation and consistency.

62. AltaGas submitted that the factors considered by the Commission in previous

proceedings continue to be relevant. Both AltaGas and AltaLink referred to previous

considerations, such as potential rate shock, rate stabilization, intergenerational equity,

regulatory burden, consistency with accounting standards, impact on credit metrics and the

treatment of any future income tax costs as no-cost capital.75 AltaLink added the stand-alone

principle as an overarching principle, which results in determining regulatory income tax on a

deemed corporation basis.76

63. The Commission also received evidence about differences in the sum of the present

discounted value of the revenue requirement and the impacts on FFO/debt, in deciding which

income tax method should be applied to the utilities.

Intergenerational equity

64. Mr. Madsen focused on the timing differences for income tax associated with overhead

costs and salvage costs, both of which are associated with capital assets. Mr. Madsen explained

that an income tax deduction is available for the full amount of the overhead costs in the year

they are incurred, whereas for ratemaking purposes, the costs are recovered over the life of the

capital asset. For ratemaking purposes, salvage costs are also collected over the life of the capital

asset, but for income tax purposes, the costs are not deductible until the year they are incurred,

which is the last year of the capital asset’s life. Under the flow-through method, these timing

differences persist, and Mr. Madsen commented that these timing differences can be significant.

He further submitted that under the FIT method, these timing differences are accounted for, and

because of this, the income tax benefit of the overhead costs and the salvage costs are allocated

to all customers over the life of the capital asset.77

65. AltaGas suggested that the findings of the Commission’s predecessor, the Public Utilities

Board, in Report E79079,78 conflict with Mr. Madsen’s argument about intergenerational equity.

It referred to the findings of the Public Utilities Board that the FIT method provides for a

potential liability rather than a real liability, with much uncertainty surrounding the probability of

payment of the future income tax component.79 AltaGas submitted that the adoption of the FIT

method would result in a higher proportion of a new capital asset’s costs being shifted to the

earlier years of its life.80 It stated that any assessment of intergenerational equity needs to be done

on the overall revenue requirement, not just the income tax component.81

66. The ATCO Utilities submitted that if utilities move from the flow-through method to the

FIT method, the resulting collection of any unfunded FIT liability will contribute to

75 Exhibit 22570-X0041, paragraph 10. 76 Exhibit 22570-X0043, paragraphs 10-13. 77 Exhibit 22570-X0557, paragraphs 37-42. 78 Report E79079: The Income Tax Component of the Utility Revenue Requirement for Alberta Utilities,

August 1, 1979. 79 Exhibit 22570-X0783, paragraph 14. 80 Exhibit 22570-X0783, paragraph 18. 81 Exhibit 22570-X0783, paragraph 18.



intergenerational inequity, because the unfunded FIT liability has accumulated over many

decades.82 The ATCO Utilities contended that any purported advantages of the FIT method in

mitigating intergenerational equity concerns are not sufficient to justify a requirement that the

utilities all convert to the FIT method.83

67. Dr. Villadsen commented that Mr. Madsen’s focus on intergenerational equity is very

narrow, because he examines what happens during two specific years of a capital asset’s life, the

first year and the last year. She submitted this focus is less relevant because of the utilities’

continuous ongoing investment in capital assets.84

68. AltaLink suggested that under the FIT method, current ratepayers would be paying higher

rates that include an income tax component that may not be payable at some uncertain future

time.85

Matching of costs and revenues

69. Mr. Madsen provided the following explanation for why the flow-through method does

not adhere to the matching of costs and revenues principle:

A. MR. MADSEN: And to be clear, though, I don't agree necessarily that there is a

better matching from a regulatory perspective. As my evidence has stated, the

Commission allows and approves the collection of numerous costs, depreciation, salvage,

overheads, a number of costs over the life of the assets from a regulatory perspective, and

yet the income taxes, from a regulatory perspective, are not collected on the same basis;

i.e., the income taxes costs are not matching the revenues that drive them.86

Cost causation and consistency

70. Mr. Madsen indicated that from an accounting perspective, the accrual method is

commonly accepted. He added that for regulatory purposes, the cost causation principle as well

as intergenerational equity drive the collection of certain items such as depreciation and salvage

over the life of a capital asset, as opposed to collecting the entire cost in one year. Mr. Madsen

submitted that the income tax impacts associated with items such as depreciation, salvage and

overhead costs should likewise be reflected in revenue requirement over the life of a capital

asset. He submitted that this would also promote the consistency principle.87

71. The CCA commented that the flow-through method requires customers to simply pay for

a cost when there is a cash outflow. It submitted that a cash outflow may not demonstrate a

causal link to the customers who drove that cost, especially when the cash outflow is determined

by the Income Tax Act, which is not intended to reflect sound ratemaking principles.88

72. AltaLink contended that there are differences between depreciation and future income

taxes. It submitted that while depreciation is based on the actual costs the utility pays for its

82 Exhibit 22570-X0746, paragraph 12. 83 Exhibit 22570-X0746, paragraph 13. 84 Exhibit 22570-X0767.01, A125. 85 Exhibit 22570-X0043, paragraph 36. 86 Transcript, Volume 7, page 1380. 87 Exhibit 22570-X0557, paragraphs 43-45. 88 Exhibit 22570-X0888, paragraph 434.



capital assets, future income taxes are a non-cash expense with uncertainty regarding their timing

and amount.89

73. Dr. Villadsen stated that regardless of which income tax method is used, customers pay

more in the early years of a capital asset’s life for the recovery of capital, and she suggested this

is so even though the service provided by the capital asset is not necessarily more valuable to

customers in those early years.90

74. The CCA described certain factors that may cause larger-use customers to decide they no

longer want to receive service from their utility. The CCA expressed its concern that if these

larger-use customers no longer receive service from their utility, the residential customers will be

left with the burden of increased costs, including future income taxes. Emphasizing the cost

causation principle, the CCA submitted that under the flow-through method, these larger-use

customers will not have paid for their share of future income taxes, even though they triggered

the costs. The CCA submitted that the continued use of the flow-through method does not result

in a fair allocation of costs among ratepayers.91 AltaLink countered that the use of the FIT

method, with its resulting increases in customer rates, will exacerbate the risk of larger-use

customers no longer wanting to receive utility service from their utility.92 Mr. Bell expressed a

similar view in response to questioning from Commission Member Lyttle at the hearing.93

Impact on rates

75. AltaLink indicated that a switch from the flow-through method to the FIT method would

result in significant increases in customer rates.94 AltaLink submitted that the impact on customer

rates should be a consideration when determining the income tax method, especially in times of a

downturn in the economy, when rate relief is most needed.95

76. AltaGas submitted that if it had to switch to the FIT method, and its total unfunded FIT

liability had to be collected from customers, it would likely result in some degree of rate shock,

and create an administrative burden to track and account for the change in the income tax

method.96

77. Mr. Madsen stated that the Commission has significant latitude to determine how the FIT

liability can be funded by customers. He indicated the liability does not need to be fully funded

in a short period of time, and the period of time over which the liability should be funded would

be specific to each utility and where that utility is within the overall life of its capital assets.

Another option noted by Mr. Madsen, though not a preferred option, would be to ignore the

accumulated FIT liability on transition, and monitor it on a go-forward basis to assess whether

collection in a future period would be possible.97

89 Exhibit 22570-X0738, paragraph 89. 90 Exhibit 22570-0767.01, A122 and A124. 91 Exhibit 22570-X0888, paragraphs 407, 409, 413. 92 Exhibit 22570-X0058, paragraphs 33-34. 93 Transcript, Volume 10, page 2171. 94 Exhibit 22570-X0043, paragraph 34. 95 Exhibit 22570-X0043, paragraph 18. 96 Exhibit 22570-X0041, paragraph 12. 97 Exhibit 22570-X0557, paragraphs 70-75.



78. AltaLink commented that the collection of future income taxes, and their treatment as

no-cost capital, introduces refinancing risk in future years when the accumulated FIT liability is

drawn down and the utility has to replace it with debt and equity, at rates that may be greater

than the period over which the future income taxes were collected.98

Consideration of the sum of the present discounted value of the revenue requirement in

determining an income tax method

79. Dr. Villadsen undertook an analysis that considered a simple illustrative model, using a

set of representative input parameters, in order to evaluate the effects of applying the flow-

through method and the FIT method to a hypothetical regulated utility.99 Based on her analysis,

Dr. Villadsen concluded that when measured purely in terms of the sum of the present value of

the revenue requirement over the full economic life cycle of a utility investment, there is no clear

advantage for customers or the utilities from either the flow-through method or the FIT

method.100

80. AltaLink submitted that present value or discounted cash flow analysis is not the best

factor to consider when deciding on an income tax method. AltaLink suggested that larger

industrial or commercial customers, who typically have higher discount rates than a utility,

prefer to keep their funds to invest in their business, rather than involuntarily paying future

income taxes.101

Impact on credit metrics

81. The CCA submitted that when deciding upon an income tax method, the Commission

should not ignore the ability of the FIT method to improve a utility’s credit metrics.102

82. Dr. Villadsen agreed that the FIT method will provide a utility with greater cash flow

early in the life of a capital asset, which can provide support for credit metrics such as FFO/debt.

However, she cautioned that later in the life of the capital asset, the situation is reversed.

Dr. Villadsen submitted that implementing the FIT method should be viewed as a long-term

commitment, and therefore it is important to consider both the short-term and long-term

consequences for credit quality, when determining an income tax method.103

Commission findings

83. The scope of the Commission’s generic proceeding on income taxes included an

exploration of whether one income tax methodology should be applied uniformly to all utilities,

or whether different methodologies should be used under different circumstances.

84. The Commission agrees that transition to the FIT method will reveal significant FIT

liabilities.104 The estimated balance at December 31, 2017, of the unfunded FIT liabilities is

98 Exhibit 22570-X0043, paragraph 46. 99 Exhibit 22570-X0170, A4. 100 Exhibit 22570-X0170, A3. 101 Exhibit 22570-X0141, paragraph 55. 102 Exhibit 22570-X0888, paragraph 473. 103 Exhibit 22570-X0170, A19. 104 See, for example, Exhibit 22570-X0557, A70.



approximately $1.4 billion.105 This balance would have to be grossed up for current income taxes

in determining rates. Using the current statutory income tax rate of 27 per cent, the result would

be approximately $1.9 billion.

85. In considering this issue, the Commission must be cognizant of revenue requirement

effects and the resulting impact on rates. Collection of approximately $1.9 billion to fund the

estimated total FIT liabilities at December 31, 2017, and the additional estimated total increase in

annual revenue requirements of approximately $200 million associated with adopting FIT, has to

be considered when deciding whether all utilities should adopt the FIT method. The Commission

must also consider the other relevant factors identified, including intergenerational equity,

matching, cost causation and consistency, and impact on credit metrics.

86. Mr. Madsen’s assessment of intergenerational equity focused on the timing differences

associated with overhead costs and salvage costs. The Commission agrees with Dr. Villadsen’s

comments that Mr. Madsen’s focus in this area was very narrow, because he focused on the

income tax deductibility of overhead costs in the first year of an asset, and the deductibility of

salvage costs in the last year.

87. The Commission considers that the two examples provided by Mr. Madsen are

representative of intergenerational issues, when considered in isolation. Under the flow-through

method, if a person is not a customer in the year a capital asset is added, but becomes a customer

in the year after the capital asset has been added, the customer will not have received the benefit

of the deduction of the overhead cost in the first year. Similarly, if a person is a customer in the

year the capital asset is added, and remains a customer until the year before the capital asset is

retired and salvage costs are incurred and deducted for income tax, that customer does not

receive the benefit of that income tax deduction.

88. However, the Commission agrees with Dr. Villadsen that these examples ignore the

continuous ongoing investment in capital assets that the utilities make. Consequently, in

Mr. Madsen’s example, the person who becomes a customer in the year after a capital asset is

added, will benefit from the deduction of the overhead costs of the capital asset that is added in

the year that person becomes a customer. Mr. Madsen’s second example, relating to salvage

costs, ignores the consideration that the customer would have benefitted from the deduction of

salvage costs that were paid out while that person was a customer.

89. The Commission finds that the issue of intergenerational equity with respect to income

tax is more complex than was represented by Mr. Madsen’s examples. One example observed by

Dr. Villadsen, with respect to the income tax reform in the U.S. in December 2017, highlighted

the potential for intergenerational equity implications. Dr. Villadsen indicated that most U.S.

utilities recover income taxes on a method that is similar to FIT. She noted that because of the

reduction in the U.S. federal income tax rate, the FIT liabilities for these U.S. companies will be

reduced, and most of these utilities now have over-collected their FIT.106 The Commission

considers this situation, which is specific to the FIT method, also raises intergenerational equity

concerns.

105 Exhibit 22570-0746, Table 1. 106 Exhibit 22570-X0767.01, A147-A148.



90. The Commission finds that the submissions of Mr. Madsen with regard to the

intergenerational issues raised by the flow-through method fail to offer convincing support for

the abandonment of this method and the adoption of the FIT method for all utilities, especially in

light of the fact that the FIT method has potential to create its own intergenerational issues.

91. Mr. Madsen submitted that the flow-though method does not adhere to the matching of

costs and revenues. He indicated that the Commission allows and approves the collection of

numerous costs such as depreciation, salvage and overheads over the life of the assets from a

regulatory perspective, yet it does not do so for income taxes. He stated that the income tax costs

are not matching the revenues that drive them. The Commission finds that Mr. Madsen’s analysis

does not account for the fact that there are differences between the utilities with respect to items

such as depreciation and salvage. For example, EPCOR recovers salvage costs over the life of

the assets that are in place subsequent to the retirement of the asset, whereas the other utilities

generally recover salvage costs during the life of the original asset.

92. The Commission considers that the actual income taxes paid to the taxation authorities

are valid income tax costs for regulatory ratemaking purposes, and these costs are matched to the

revenues that drive them. While the liability for future income taxes exists, the measurement of

that liability, as reflected on a utility’s balance sheet, is done at a certain point in time, based on a

number of assumptions. It is assumed that no other capital assets will be added and that future

depreciation rates, capital cost allowance rates and statutory income tax rates, among others, will

remain constant. These assumptions and correspondingly, the FIT liability, change from year to

year. This adds much uncertainty as to whether the FIT expense is an accurate representation of

the actual income taxes the utility will pay to the taxation authorities in subsequent years. The

situation where U.S. utilities operating under FIT recently experienced a reduction in the

statutory tax rate highlights this point. The Commission considers this uncertainty supports

arguments against the FIT method being adopted as the standard income tax method for revenue

requirement purposes.

93. Mr. Madsen commented that in order to promote the consistency principle, depreciation,

salvage and overhead costs are collected over the life of a capital asset, and so should the income

tax impacts associated with these items. For regulatory purposes, overhead costs form part of the

cost of a capital asset, and are therefore recovered through depreciation. Salvage costs are linked

to the cost of retiring a capital asset. The Commission considers that, even in isolation, the actual

income tax associated with a capital asset over its life cannot be determined without a number of

assumptions, as the Commission has commented on above. When this is combined with the fact

that the utilities are adding capital assets on an annual basis, this income tax determination is

made even more difficult.

94. A depreciation rate is applied to a class of capital assets in order to ensure that the cost of

the capital assets is recovered. The cost of the capital assets is known. Income taxes, as their

description suggests, are associated with income, and the income over the life of any particular

capital asset is unknown. The reasons for collecting depreciation over the life of a capital asset

for regulatory purposes are well known and are grounded in the collection of the original cost of

the capital asset. The collection of income taxes is not limited to the original cost of any

particular capital asset, but instead involves additional factors such as income tax rates and the

availability of income tax deductions. In consideration of all of the above, the Commission finds

that there is no need for consistent treatment between the collection of depreciation, salvage

costs and overheads, and the uncertain income tax impacts associated with them.



95. For all these reasons, the Commission finds that Mr. Madsen’s recommendation for the

use of the FIT method is not supported. Given this finding, and the Commission’s understanding

that the adoption of the FIT method would result in significant cost implications for customers,

the Commission will not require every utility to uniformly adopt the FIT method.

The Commission finds that the use of the flow-through method is acceptable, and should

continue to be used as the default method.

96. The Commission does not consider that the foregoing should prevent a utility from

applying to adopt the FIT method in a future rate-related proceeding. The onus will be on the

applicant proposing FIT in a future rate-related proceeding to satisfy the Commission that the

specific circumstances warrant a change to the FIT method.

5.2 Claiming maximum allowable income tax deductions when forecasting income

tax expense

97. One issue with regard to determining cash income taxes is whether the utility should

claim the maximum allowable deductions for income tax purposes, even if it results in taxable

income less than zero. The CCA endorsed AltaLink’s policy of not triggering taxable income of

less than zero for forecast purposes, and it submitted that all the utilities should follow this

practice.107

98. AltaGas submitted that any requirement to maximize income tax deductions and carry-

back income tax losses, instead of carrying them forward, will result in lost savings in the

situation where income tax rates are increasing.108 It indicated that income tax losses, when

carried forward, expire after 20 years, whereas discretionary deductions such as the claiming of

capital cost allowance, can be carried forward indefinitely.109 AltaGas submitted the utilities

should be given flexibility to claim the proper level of discretionary income tax deductions given

the circumstances as this would give them the best opportunity to manage their income tax

portfolios and realize available cost savings.110

Commission findings

99. The Commission finds that because of the finite life of income tax loss carryforwards, as

opposed to the indefinite life of deductions such as capital cost allowance, the conservative

practice would be for utilities not to forecast income tax losses, but instead, forecast the use of

discretionary deductions such as capital cost allowance in order to reduce forecast taxable

income to zero. Accordingly, the Commission directs the utilities, when forecasting income

taxes, to only claim allowable deductions that will reduce the taxable income to a maximum of

zero.

5.3 Reporting of future income tax liabilities

100. Mr. Madsen recommended that the Commission require the utilities to quantify their

accumulated and unreported FIT liability under International Financial Reporting Standards each

year, and report this information each year as part of their Rule 005: Annual Reporting

Requirements of Financial and Operational Results. He also recommended that each utility

107 Exhibit 22570-X0557, paragraphs 100-103. 108 Exhibit 22570-X0041, paragraph 17. 109 Exhibit 22570-X0041, paragraph 20. 110 Exhibit 22570-X0127, paragraph 28.



should propose a method to fund the FIT liability as part of their next cost-of-service application

or PBR filing.111 Mr. Madsen and the CCA submitted that this information should be reported,

even if the Commission continues to approve the use of the flow-through method.112

101. With respect to the CCA’s recommendation for utilities to report their unfunded FIT

liability, AltaGas noted that this information is currently reported as part of its annual audited

financial statements.113 Both AltaGas and the ATCO Utilities submitted that any amendments to

Rule 005 are outside the scope of this GCOC proceeding. They indicated that reporting the

unfunded FIT liability is not warranted as part of Rule 005, because this liability has no bearing

on the utility’s financial performance. AltaGas and the ATCO Utilities proposed that the CCA’s

recommendation be dismissed.114

Commission findings

102. The Commission agrees with AltaGas and the ATCO Utilities that reporting the unfunded

FIT liability would have no bearing on their financial performance. However, given the

magnitude of the unfunded FIT balances that were forecast as of December 31, 2017, and the

Commission’s consideration that the calculation and reporting of this balance on an annual basis

would not require a significant amount of effort, the Commission directs the ATCO Utilities,

FortisAlberta, AltaGas and AltaLink to include their unfunded FIT liability balance each year as

part of their Rule 005 reports, beginning with the Rule 005 report for 2018, that will be

submitted in 2019. The information provided should consist of the unfunded FIT liability for the

year being reported, as well as the previous year, and the resulting difference. This information

may assist the Commission in assessing the level of potential credit metric relief that may be

available if a utility were to apply to adopt the FIT method.

5.4 Income tax deferral accounts

Criteria for establishing deferral accounts

103. The ATCO Utilities recommended that any deferral accounts for income taxes be

established in accordance with the criteria the Commission has previously applied. The ATCO

Utilities described these criteria as being (1) the materiality of the forecast amount;

(2) uncertainty regarding the accuracy of the forecast amount; (3) uncertainty regarding the

ability of the utility to forecast the amount; (4) whether or not the factors affecting the forecast

are typically beyond the utility’s control; and (5) whether or not the utility is typically at risk

with respect to the forecast amount.115

104. Mr. Bell recommended that the use of deferral accounts be minimized. He submitted that

deferral accounts transfer risk from the utility to customers. Mr. Bell stated that a deferral

account should only be allowed if it satisfies the criteria established by the Commission for

Y factor treatment. These criteria include (1) the amounts are attributable to events outside

management’s control; (2) have significant influence on the operation of the company; (3) do not

have a significant influence on the inflation factor (I factor) in the PBR formula; (4) have been

111 Exhibit 22570-X0557, paragraph 76. 112 Exhibit 22570-X0557, paragraph 23. Exhibit 22570-X0888, paragraph 402. 113 Exhibit 22570-X0783, paragraph 20. 114 Exhibit 22570-X0918, paragraph 270. 115 Exhibit 22570-X0044, paragraph 11.



prudently incurred; and (5) are of a recurring nature with the potential for a high level of

variability.116

Deferral account for statutory income tax rates and capital cost allowance rates

105. AltaLink submitted that a deferral account should be used for statutory income tax rate

changes as well as changes to capital cost allowance rates.117 It added that changes in these rates

could be material and are beyond the control of a utility.118 AltaGas and Mr. Madsen agreed with

establishing deferral accounts for changes in statutory income tax rates and changes in capital

cost allowance rates.119 Mr. Bell commented that changes in statutory income tax rates clearly

qualify for deferral account treatment.120

106. In accordance with the criteria for deferral accounts they put forward, the ATCO Utilities

suggested that a deferral account for changes in statutory income tax rates be established.121

107. Mr. Madsen and AltaGas commented on the Commission’s finding in Decision 2012-

237122 that changes in statutory income tax rates impact the entire economy and should be

captured by the I factor for the PBR utilities. Mr. Madsen and AltaGas stated that there is a lag in

the impact of a change in the statutory income tax rate on the I factor.123 AltaGas contended that

there is no direct causal link between changes in statutory income tax rates and inflation.124

Mr. Madsen submitted that changes such as increasing revenue requirements and income taxes,

the potential for changes in governments and the significant income tax changes implemented in

recent years are outside the control of the utilities, and this supports the use of deferral accounts

for changes in statutory income tax rates.125

Deferral account for temporary differences and income tax reassessments

108. Mr. Madsen supported the continued inclusion of temporary differences for income tax

within the currently established direct assigned capital deferral accounts for the transmission

utilities that operate under cost of service.126 Mr. Madsen suggested that elective income tax

planning strategies, such as the use of rolling starts for lengthy projects, should be the subject of

a deferral account, unless the Commission can incorporate the effects of such income tax

planning strategies in a way that allows customers to share in the future benefits.127

109. Mr. Madsen submitted that in the case of utilities that have a history of poor forecasting

accuracy with regard to temporary income tax differences that are not subject to an existing

116 Exhibit 22570-X0559, A25. 117 Exhibit 22570-X0043, paragraph 5. 118 Exhibit 22570-X0043, paragraph 20. 119 Exhibit 22570-X0783, paragraph 35. Exhibit 22570-X0557, paragraph 85. 120 Exhibit 22570-X0559, A25. 121 Exhibit 22570-X0044, paragraph 12. In Exhibit 22570-X0171, paragraph 21, the ATCO Utilities provided more

details about how any changes in statutory income tax rates would be addressed as part of an adjustment to

revenue. 122 Decision 2012-237: Rate Regulation Initiative, Distribution Performance-Based Regulation, Proceeding 566,

Application 1606029-1, September 12, 2012. 123 Exhibit 22570-X0557, paragraph 86. Exhibit 22570-X0041, paragraph 26. 124 Exhibit 22570-X0041, paragraph 26. 125 Exhibit 22570-X0557, paragraph 86. 126 Exhibit 22570-X0557, paragraph 90. 127 Exhibit 22570-X0557, paragraph 106.



deferral account, the Commission should determine whether a reserve or deferral account is

necessary. He indicated that a deferral account should be utilized until such time as the utility

can demonstrate a clear ability to properly forecast its income tax expense. Mr. Madsen

suggested the Commission make these assessments on a case-by-case basis.128

110. The ATCO Utilities pointed out that for utilities under PBR, there is no forecasting

accuracy to assess with respect to income taxes. They added this is because the income tax

expense is determined at the beginning of the PBR term, and it is indexed for each subsequent

year. The ATCO Utilities contended that income tax expense calculated under the flow-through

method comprises approximately three per cent of the total revenue requirement, and this small

percentage should be considered when deciding whether deferral accounts for income taxes are

required.129

111. AltaGas disagreed with Mr. Madsen’s recommendation for the inclusion of a deferral

account for temporary differences, and recommended the elimination of the Y factor for income

tax timing differences for the 2018-2022 PBR term. It submitted that the continued true-up of

only the income tax effect of temporary timing differences through a Y factor would amplify the

under or over recovery of a utility’s capital revenue requirement.130 It submitted that under the

2018-2022 PBR term, the risks of capital investment have shifted to the utilities and any deferral

account for temporary differences would be inconsistent with the incentive properties for

PBR under the K-bar funding mechanism.131

112. AltaLink submitted that a deferral account for the deemed regulatory tax expense is not

required, as long as there are applicable deferral accounts for the other non-income-tax revenue

requirement components that cause differences between forecast and actual income tax, such as

the deferral account for direct assigned capital project costs.132

113. FortisAlberta proposed that an income tax deferral account is justified and necessary.133

FortisAlberta stated that deferral account treatment is a means for it to recover income tax

amounts subject to reassessment by the Canada Revenue Agency (CRA). It noted that it has

never been audited by the CRA, and until such an audit occurs, there is uncertainty with respect

to the income tax deductions claimed.134 FortisAlberta explained that a deferral account also

provides for recovery of temporary income tax differences, such as contributions that it is

required to pay to the Alberta Electric System Operator (AESO). FortisAlberta stated that these

contributions are subject to significant variability and will materially impact income tax

expense.135

114. The ATCO Utilities submitted that, when assessed against its recommended criteria for

establishing deferral accounts, the Y factor dealing with income tax reassessments and temporary

128 Exhibit 22570-X0557, paragraph 89. 129 Exhibit 22570-X0746, paragraph 24. 130 Exhibit 22570-X0783, paragraphs 31-34. 131 Exhibit 22570-X0783, paragraph 31. 132 Exhibit 22570-X0043, paragraph 71. 133 Exhibit 22570-X0039, paragraph 34. 134 Exhibit 22570-X0039, paragraph 32. 135 Exhibit 22570-X0039, paragraph 33.



income tax differences established under the first PBR term is not required for the 2018-2022

PBR term.136

Commission findings on deferral account proposals

Criteria for establishing deferral accounts

115. The ATCO Utilities recommended that any deferral accounts for income taxes be

established in accordance with the five criteria previously established by the Commission.

Mr. Bell stated that deferral accounts should be established based on the five criteria set out by

the Commission in its decision on the 2018-2022 PBR term.137 The Commission will use the

criteria referenced by Mr. Bell in assessing what deferral accounts, if any, should be established

for income taxes for the distribution utilities.

116. The Commission finds that the five criteria listed by the ATCO Utilities should form the

basis upon which any deferral accounts for income taxes for the transmission utilities should be

decided. In addition, the Commission considers that the symmetry factor detailed in paragraphs

71-74 of Decision 2010-189138 should also be considered, as “symmetry must exist between costs

and benefits for both the Company and its customers.”139 However, the Commission will not make

any specific findings with respect to income tax deferral accounts for the transmission utilities in

this decision. The Commission considers that determinations with respect to tax deferral

accounts for the transmission utilities are best made on the basis of a utility’s specific

circumstances and on a case-by-case basis, and considering the criteria articulated in this

decision.

117. With respect to the distribution utilities, the Commission makes the following findings in

relation to deferral accounts in the case of (1) statutory income tax rates and capital cost

allowance rates; (2) temporary differences and income tax reassessments; and (3) other deferral

accounts.

Deferral account for statutory income tax rates and capital cost allowance rates

118. The Commission notes that the five criteria cited by Mr. Bell were those established by

the Commission for the identification of eligible Y factor costs.140 The third criterion for eligible

Y factor treatment requires that costs should not have a significant influence on the inflation

factor in the PBR formula. The Commission decided in Decision 2012-237 that major changes to

the calculation of income tax payments, such as a change in income tax rates, should impact the

entire economy and, as such, should be captured by the I factor. The Commission stated that due

to the infrequent nature of such changes, it was not necessary to establish a Y factor account for

changes in statutory income tax rates.141

136 Exhibit 22570-X0044, paragraph 17. 137 Decision 2012-237, paragraph 631. 138 Decision 2010-189: ATCO Utilities, Pension Common Matters, Proceeding 226, Application 1605254-1,

April 30, 2010. 139 Decision 2010-189, paragraph 73, quoting page 148 from Decision 2000-9: Canadian Western Natural Gas

Company Limited, 1997 Return on Common Equity and Capital Structure, and 1998 General Rate Application,

Applications 980413 and 980421, March 2, 2000. 140 Decision 2012-237, paragraph 631. 141 Decision 2012-237, paragraph 711.



119. In this proceeding, Mr. Madsen and AltaGas, as well as Dr. Carpenter,142 argued that there

is a lag to the impact of any changes to the statutory income tax rates being reflected in the

I factor. The ATCO Utilities and AltaGas recommended that a deferral account be established

for changes in statutory income tax rate changes. Mr. Bell and Mr. Madsen agreed.

120. The Commission is not persuaded by the submissions of parties that deferral account

treatment for changes in statutory income tax rates is necessary to account for any lag in the

I factor. The Commission maintains its finding on this issue from Decision 2012-237, that it is

not necessary to establish a Y factor account for changes in statutory income tax rates. By

extension, this finding extends to any changes in capital cost allowance rates.

121. The Commission addressed concerns with respect to the lagged approach to the I factor in

Decision 2012-237 and concluded as follows:

243. The main difference between the two methods is that the approach preferred by

the ATCO companies and Fortis ensures that the impact of actual inflation in any given

year is reconciled soon after the year‘s end, while the alternative approach of using the

actual inflation from the previous year involves a certain lag for such a true-up to occur.

In this proceeding, parties’ concerns with the lagged approach seemed to be centered on

the fact that the lag between the inflation index used in the PBR formula and the actual

inflation experienced in the economy would expose the companies to windfall gains

or losses, although these would be transitory. [emphasis added]

244. The Commission considers that if inflation is higher in some years and lower in

other years, as appears to be the general case in the economy, then using the most recent

historical inflation rate will average out the effect of any regulatory lag over the PBR

period. Indeed, as ATCO Gas observed in its argument, in the absence of a true-up, the

I factor in 2009 would be higher than actual inflation. The opposite would have occurred

in 2010, where the I factor without the true-up would be lower than actual inflation. As

such, inflation will tend to balance out over the PBR term, obviating the need to

true-up the I factor through a separate regulatory proceeding. [emphasis added]

…

248. In light of these considerations, the Commission finds that the lagged

approach currently used by ENMAX and proposed by AltaGas and EPCOR in this

proceeding represents a better alternative as compared to the forecast and true-up

method proposed by the ATCO companies and Fortis.... The Commission considers

that this approach will provide a fair balance between accuracy and regulatory

efficiency and will make the companies‘ PBR plans more transparent and simple to

understand thereby furthering the objectives of the third Commission PBR

principle. [emphasis added]

122. The Commission therefore denies the request of the ATCO Utilities, AltaGas, Mr. Bell

and Mr. Madsen that deferral accounts be established for the distribution utilities to account for

any changes in statutory income tax rates and capital cost allowance rates.

142 Exhibit 22570-X0186, A72.



Deferral account for temporary differences and income tax reassessments

123. Mr. Madsen recommended that a deferral account be established for all temporary

differences. AltaGas argued against this recommendation. The Commission finds that the scope

of this deferral account is too broad, and is not in accordance with the applicable criteria. The

Commission considers that all costs that give rise to temporary differences are not outside the

control of the utility’s management, and not all of these costs are material. Consequently,

Mr. Madsen’s recommendation for a deferral account for all temporary differences is denied.

124. FortisAlberta currently has a deferral account for the income tax expense impact of

reassessments made by the CRA in respect of income tax deductions that FortisAlberta has

taken. With respect to this deferral account, the Commission finds that this is not beyond

management’s control, but is rather a safeguard against the actions taken by the management of

FortisAlberta in deciding what income tax deductions should be taken by the company.

FortisAlberta noted that maximizing allowable deductions for income tax, including

Canderel/Rainbow pipeline expenditures, and capitalized overhead expenditures has allowed it to

remain in a non-tax paying position.143 FortisAlberta indicated that AESO contributions are a

material deduction that it relies on to maintain a non-tax-paying position.144

125. The Commission considers that the decision to take these income tax deductions was, and

is, within the control of the management of FortisAlberta. FortisAlberta was not required to take

these income tax deductions, and if it had not taken the deductions, and ended up in a taxable

position, then the resulting income tax expense would have formed part of the going-in rates for

PBR, and customers would have been required to pay the resulting income taxes.

126. The Commission is aware that FortisAlberta had a CRA reassessment deferral account in

place for 2012.145 This was part of a negotiated settlement, and therefore it may have been part of

the gives and takes associated with negotiated settlements. The Commission notes that it had

previously denied FortisAlberta’s request to establish a CRA reassessment deferral account.146

127. Based on the foregoing, the Commission denies the request of FortisAlberta that a CRA

reassessment deferral account, in the form of a Y factor, be established for it for the 2018-2022

PBR term. However, in acknowledgement that a deferral account was in place for 2012-2017,

should FortisAlberta be reassessed in relation to income tax expense for this period, it may bring

this matter forward for consideration by the Commission.

Other deferral accounts

128. FortisAlberta currently has a deferral account in case it becomes subject to income tax

over the term of the PBR plan. With respect to this deferral account, the Commission considers

that this is linked to the deductibility of the Canderel/Rainbow pipeline expenditures, disposal

costs and AESO contributions, which the Commission finds are within the control of

management. Consequently, this deferral account does not meet the criteria for being outside the

143 Exhibit 22570-X0039, paragraph 28. 144 Exhibit 22570-X0039, paragraph 33. 145 Decision 2012-108: FortisAlberta Inc., Application for Approval of a Negotiated Settlement Agreement in

respect of 2012 Phase I Distribution Tariff Application, Proceeding 1147, Application 1607159-1, April 18,

2012, paragraph 52. 146 Decision 2010-309: FortisAlberta Inc. 2010-2011 Distribution Tariff – Phase I, Proceeding 212, Application

1605170-1, July 6, 2010, paragraph 307.



control of management, and the Commission therefore denies FortisAlberta’s request to establish

a deferral account for income taxes, in case the company becomes taxable over the term of the

PBR plan.

129. The ATCO Utilities undertook an assessment of their currently established income tax

deferrals, and whether they met the criteria previously applied by the Commission for the

establishment of deferral accounts.147 While these criteria are not exactly the same as the criteria

established by the Commission for Y factor treatment, two of the criteria, being the materiality of

the amount, and whether it is outside management’s control, are the same.

130. Based on their assessment that the costs were immaterial and their assessment that

management can control them, the ATCO Utilities recommended that the deferral account

currently in place for the distribution utilities regarding income tax deductible capital costs is not

necessary and should be eliminated.148 The ATCO Utilities also submitted that the deferral

account currently in place with respect to the deductibility of deferral accounts for income tax

purposes relates to deferral accounts that are in place for non-income tax deferrals. They

submitted that as long as these non-tax-related deferral accounts, such as ATCO Gas’s weather

deferral account, remain in place, it is fair that the income tax impact associated with these non-

tax-related deferral accounts remain in place.

131. The Commission agrees with the assessment of the ATCO Utilities in support of their

recommendation that the deferral account they currently have in place for income tax deductible

capital costs is not necessary. These costs are not attributable to events outside the control of

management. The utility’s management will be responsible for monitoring changes to the income

tax legislation and keeping apprised of any relevant court cases that may help identify any

potential deductions that could be claimed. Management has to be satisfied that the deductions

are justified and will stand up to any subsequent scrutiny by the taxation authorities. Based on

this, the Commission finds that it is not necessary for the distribution utilities to establish a

deferral account for any income tax deductible capital costs. The Commission also agrees with

the ATCO Utilities that as long as any non-tax-related deferral accounts remain in place for the

distribution utilities, the income tax aspect of these deferrals is to remain in place as well.

5.5 PBR implications

Altering of income tax assumptions and practices during the PBR term, or on rebasing

132. AltaGas submitted that a utility should be able to alter its income tax assumptions and

practices during the PBR term, if incentives regarding income taxes are built into the regulatory

process. This would permit the utility to optimize its income tax position and mitigate potential

income tax liabilities. Noting that income tax addbacks and deductions are derived from the

accounting records, AltaGas suggested that any examination of changes in accounting

assumptions and practices during the PBR term should also include a consideration of the impact

on income taxes.149

133. Mr. Madsen recommended that the FIT method be implemented within going-in rates for

the PBR utilities. He commented that the incentive for PBR utilities to claim maximum income

147 Exhibit 22270-X0171, PDF pages 5-8. 148 Exhibit 22270-X0171, PDF pages 5-8. 149 Exhibit 22570-X0041, paragraphs 28-29.



tax deductions under the flow-through method will result in these deductions not being available

for customers in the future, and therefore “is not a proper incentive and not in the public

interest.”150 Mr. Madsen indicated that the use of the FIT method would partially address all the

income tax irregularities that a utility can benefit from under PBR. He stated that under the FIT

method, income tax expense from year to year would only fluctuate to the extent that net income

fluctuates, unlike the current volatility of income tax levels for the PBR utilities.151

Commission findings

134. The Commission notes that the income tax expense component of the going-in rates for

the 2018-2022 PBR term have been treated as a placeholder, pending the outcome of this GCOC

proceeding.152 Given that the Commission will not be directing a change to the income tax

methodology for the taxable distribution utilities and has not adopted an all-inclusive Y factor for

the treatment of income tax, few revisions to the income tax expense placeholders will be

required. However, as a result of the Commission’s decision to eliminate AltaGas’ Y factor for

tax-timing differences, AltaGas’s 2018 base K-bar calculation will need to be revisited.153

135. The Commission notes that adjustments will be made to the distribution utilities’ going-in

PBR rates in future proceedings. For example, adjustments to going-in rates will be required to

reflect 2017 approved capital tracker amounts and to account for any approved depreciation

changes. The Commission directs AltaGas to revise the calculation of its base K-bar to

incorporate the findings in this decision as part of the next proceeding addressing adjustments to

AltaGas’s going-in PBR rates. To the extent that ATCO Gas, ATCO Electric or FortisAlberta

consider that this decision impacts the calculation of the income tax expense included in 2018

going-in rates, this may similarly be addressed in the next proceeding considering any required

adjustments to their respective going-in PBR rates.

136. The Commission considers that, as a result of eliminating the majority of Y factor

treatment for income tax related matters, it is incumbent upon a taxable utility under PBR to

notify the Commission of any changes to its tax policy or other changes that may result in

changes to the utility’s taxable income and/or income tax expense. The Commission reminds the

taxable distribution utilities that the required attestation certificates filed in the annual PBR rate

adjustment filings must identify and describe any changes in accounting methods, including

assumptions respecting capitalization of labour and overhead and associated impacts.154

6 Relevant changes in global economic and Canadian capital market conditions

since the 2016 GCOC decision

137. Consistent with its practice in past GCOC decisions, in this section the Commission

considers prevailing economic and market conditions in its determination of a fair approved

ROE and approved deemed equity ratios.

150 Exhibit 22570-X0557, paragraph 235. 151 Exhibit 22570-X0557, paragraphs 232-235. 152 Decision 22394-D01-2018: Rebasing for the 2018-2022 PBR Plans for Alberta Electric and Gas Distribution

Utilities, First Compliance Filing, Proceeding 22394, February 5, 2018, Sections 6.3.1 and 6.3.2. 153 In Decision 22394-D01-2018, the Commission approved AltaGas’ base K-bar calculation, which excluded any

tax implications given AltaGas’s Y factor, on a placeholder basis. 154 Decision 2012-237, paragraph 862.



138. In the 2016 GCOC decision, the Commission concluded that the “global and Canadian

economic and capital market conditions were different from the conditions that existed during

the global financial crisis of 2008-2009,” but lingering effects of the global financial crisis

continued.155 The Commission was presented with forecasts that indicated a continued lacklustre

performance of the Canadian economy in 2016, with gross domestic product (GDP) growth and

inflation forecasts below historical averages, but with some amount of recovery expected by the

end of 2017. At that time, the Commission was persuaded that interest rates were likely to rise in

2017, but was uncertain about the speed and magnitude of the expected increase.

139. In the present GCOC proceeding, there was general consensus among witnesses that

market conditions have improved since the time of the 2016 GCOC decision, as demonstrated by

central banks raising policy interest rates, monetary stimulus programs in the U.S. and Europe

continuing to unwind, a moderate recovery in oil prices and the strengthening of the Canadian

dollar (CAD), among other indicators.156

140. A summary of the evidence and submissions on aspects of global and Canadian

macroeconomic conditions and how these should influence the Commission’s determination of a

fair approved ROE and deemed equity ratios is presented below.

6.1 Macroeconomic conditions

141. Mr. Buttke, Mr. Hevert and Dr. Cleary agreed that there has been generally positive and

strengthening economic growth globally and in North America since the 2016 GCOC decision.157

Mr. Hevert cited a March 8, 2018 speech by one of the Bank of Canada’s deputy governors that

referred to this as “geosynchronous growth.”158 Mr. Buttke referred to a report by the

International Monetary Fund (IMF) indicating that a global cyclical recovery is underway, and

that world GDP growth is expected to rise from 3.2 per cent in 2016 (the weakest annual growth

since the financial crisis) to 3.6 per cent in 2017 and 3.7 per cent in 2018.159

142. In his evidence, Mr. Thygesen cast doubt on historical and projected global and national

economic growth.160 However, during the oral hearing he agreed that the data shows positive

GDP growth since the 2016 GCOC decision, and that this growth is forecast to continue.161

143. Mr. Coyne considered global economic growth to be about the same as during the 2016

GCOC proceeding, saying that global “GDP is not growing at breakout levels, but we haven’t

experienced another recession.”162

144. Regarding the U.S. economy, witnesses generally agreed that it has been on a path of

mostly solid growth and job creation for the last few years.163 Mr. Coyne cited recent remarks by

155 Decision 20622-D01-2016, paragraph 81. 156 Transcript, Volume 3, pages 560-562. Transcript, Volume 5, pages 905-907. Transcript, Volume 6, pages 1154-

1156. Transcript, Volume 10, pages 2071-2072. 157 Transcript, Volume 3, page 560. Transcript, Volume 6, pages 1154-1155. Transcript, Volume 10, pages 2071-

2072. 158 Transcript, Volume 6, page 1155. Exhibit 22570-X0823, PDF page 3. 159 Exhibit 22570-X0179, PDF page 13. 160 Exhibit 22570-X0551, PDF page 35. 161 Transcript, Volume 8, pages 1657-1659. 162 Transcript, Volume 5, page 906.



the U.S. Federal Reserve System (the Fed) that economic activity in the U.S. “has been rising

moderately [in 2017] and is expected to continue its moderate pace of expansion over the next

three years.”164 Mr. Buttke pointed to the Bank of Canada’s October 2017 Monetary Policy

Report (MPR) that stated the U.S. economy is projected to expand at a moderate pace; about two

per cent on average over 2017 to 2019.165

145. With regard to the Canadian and Alberta economies, Dr. Cleary stated that Canadian

economic growth exceeded expectations during 2017, and both Canada and Alberta are expected

to experience more moderate but solid GDP growth going forward.166 Dr. Cleary referred to the

Bank of Canada’s October 2017 MPR, which predicts real GDP growth of 3.1 per cent in 2017,

followed by growth rates of 2.1 per cent in 2018 and 1.5 per cent in 2019.167 Dr. Cleary

concluded that “the Canadian and Alberta economies are expected to grow at subdued, but

healthy levels in the intermediate term.”168 The CCA stated that while the forecasts for national

GDP growth are positive, the trend is a declining or slowing one: 3.0 per cent in 2017, 2.2 per

cent in 2018 and 1.6 per cent in 2019.169

146. The utilities’ witnesses generally agreed with the GDP values presented by Dr. Cleary

and referenced similar growth figures.170 For example, Mr. Buttke drew the Commission’s

attention to a March 8, 2018 speech in which one of the Bank of Canada’s deputy governors

stated “the Canadian economy is progressing well. Following a decade of many setbacks, 2017

was a year of robust economic growth – 3 per cent for the year as a whole.”171

147. Mr. Coyne,172 Mr. Buttke,173 Dr. Cleary174 and Mr. Thygesen175 reminded the Commission

that future global and national growth is uncertain. Some witnesses referred to the presence of

market volatility as indicative of this global uncertainty.176 All witnesses acknowledged that there

was substantial uncertainty around U.S. trade policy, notably the current renegotiation of the

North American Free Trade Agreement (NAFTA), and its potential impact on Canadian and

North American growth projections.177

148. For Alberta, Mr. Buttke and Dr. Cleary both explained that projections for economic

growth have been significantly upgraded since the 2016 GCOC decision, when Alberta was

immediately struggling with the collapse in oil prices and negative GDP growth of 3.6 per cent

163 Exhibit 22570-X0179, PDF page 18. Exhibit 22570-X0153.01, PDF page 96. Exhibit 22570-X0131, PDF

page 19. Exhibit 22570-X0562.01, PDF pages 17-18. 164 Exhibit 22570-X0131, PDF page 19. 165 Exhibit 22570-X0179, PDF page 20. 166 Exhibit 22570-X0562.01, PDF page 5. 167 Exhibit 22570-X0562.01, PDF page 19. 168 Exhibit 22570-X0562.01, PDF page 77. 169 Exhibit 22570-X0888, paragraph 31. 170 Transcript, Volume 3, page 563. Exhibit 22570-X0179, PDF page 13. Transcript, Volume 5, page 908.

Transcript, Volume 6, pages 1157-1158. 171 Transcript, Volume 4, page 633. Exhibit 22570-X0823, PDF page 3. 172 Exhibit-22570-X0131, PDF page 24. 173 Exhibit 22570-X0179, PDF page 15. 174 Exhibit 22570-X0562.01, PDF page 21. 175 Exhibit 22570-X0551, paragraph 43. 176 Exhibit-22570-X0131, PDF page 24. Exhibit 22570-X0179, PDF page 16. 177 Exhibit 22570-X0153.01, PDF pages 36-42. Exhibit 22570-X0179, PDF page 6. Transcript, Volume 3, page

562. Transcript, Volume 5, page 907. Transcript, Volume 6, page 1155. Transcript, Volume 10, page 2071.



in 2016. Mr. Buttke cited Bloomberg data, which forecast Alberta’s GDP to be 4.1 per cent in

2017, 2.5 per cent in 2018 and 2.0 per cent in 2019.178

149. Notwithstanding Mr. Buttke’s general view of Alberta’s economic growth projections, he

tempered this view by noting that global oil prices have risen since the 2016 GCOC decision, in

that the benchmark price known as West Texas Intermediate (WTI) moved to over $60 per barrel

in February 2018, from just below $50 per barrel in May 2016.179 However, Alberta’s producers

are expected to miss out on much of that improvement since the Alberta benchmark price known

as Western Canada Select (WCS) fell from around $40 per barrel in early 2017 to around $35 per

barrel in February 2018. As a result, the WTI-WCS discount had widened to nearly $30 per

barrel in early 2018, compared to $10-$24 per barrel in 2017, thus significantly reducing the

benefit to Albertan producers of higher WTI prices.180

150. In the period leading up to the 2016 GCOC decision, the CAD weakened significantly

against the U.S. dollar (USD). The CAD/USD exchange rate was not a concern among witnesses

during this proceeding. Mr. Buttke explained that in late 2016 and through most of 2017, the

CAD strengthened relative to the USD, with the CAD/USD exchange rate settling at around

$0.80 USD. Mr. Buttke also referred to Bloomberg’s panel of economists, which “expect the

CAD/USD exchange to stabilize around current levels and strengthen marginally to 83 cents

over the next few years.”181 All other witnesses acknowledged and confirmed this expected

stabilization.182

151. AltaLink, EPCOR and Fortis referred to the comments of Mr. Hevert, stating that

although growth projections for Alberta’s GDP, oil prices and the CAD/USD exchange rate may

be relevant for firms considering their own growth projections, these should not be the focus of

the Commission’s analysis in this proceeding. Mr. Hevert explained that the Commission should

focus on the broader North American market in which capital is raised in order to capture the

true opportunity cost of capital.183

6.2 Inflation

152. Dr. Cleary explained that Canadian inflation from 1992 to 2016 averaged 1.81 per cent,

with a median of 1.75 per cent. This is within the Bank of Canada’s one to three per cent target

range, established since the policy’s adoption in 1991, and in line with its target rate of two per

cent.184

153. Dr. Cleary cited the Bank of Canada’s prediction in its October 2017 MPR that inflation

will remain below the bank’s target rates of 1.5 per cent in 2017 and 1.7 per cent in 2018, before

178 Exhibit 22570-X0562.01, PDF page 29. Exhibit 22570-X0749, PDF page 104. 179 Decision 20622-D01-2016, paragraph 42. 180 Exhibit 22570-X0749, PDF page 103. 181 Exhibit 22570-X0179, PDF pages 42-43. 182 Transcript, Volume 3, page 140. Transcript, Volume 5, page 907. Transcript, Volume 6, page 1155. Transcript,

Volume 10, page 2071. 183 Exhibit 22570-X890.01, paragraph 21. 184 Exhibit 22570-X0562.01, PDF page 7.



increasing to 2.1 per cent in 2019. Dr. Cleary noted that these predictions were in line with those

of the Consensus Economics forecast and the IMF.185

154. Mr. Buttke and Mr. Hevert provided similar evidence pointing to rising inflation since

2016, and a broad expectation that inflation will continue to rise modestly toward two per cent in

the U.S. and Canada. In contrast, Mr. Thygesen argued that inflation is low and falling.186

6.3 Interest rate environment

155. At the close of record for the 2016 GCOC proceeding on June 29, 2016, the U.S. federal

funds rate and the Bank of Canada’s overnight interest rate, both short-term policy interest rates,

were at 0.5 per cent.187 Since the 2016 GCOC decision was issued on October 7, 2016, the Fed

has raised the target for the U.S. federal funds rate five times, to 1.75 per cent as of March 21,

2018.188 The Bank of Canada raised its overnight interest rate three times over the same period, to

1.25 per cent as of January 17, 2018.189

156. Figure 1 below depicts the yield curves for Government of Canada (GOC) and U.S.

government bonds as of March 26, 2018. In past GCOC proceedings, witnesses explained that

monetary policy works at the short end of the yield curve via the overnight rate, and its influence

weakens as the maturity of the bond increases. Therefore, normal yields on long-term GOC

bonds are not as affected by current monetary policy as short-term interest rates are.190

Figure 1 Yield curves for GOC and U.S. government bonds as of March 26, 2018191

185 Exhibit 22570-X0562.01, PDF page 20. 186 Exhibit 22570-X0551, PDF page 37. 187 Decision 20622-D01-2016, paragraph 50. 188 Exhibit 22570-X0153.01, PDF page 28 indicates increases on these dates: December 14, 2016, March 15, 2017,

and June 14, 2017. Exhibit 22570-X0767.01, PDF page 29 indicates an increase on December 13, 2017. Exhibit

22570-X0851, PDF page 1 indicates an increase on March 21, 2018, to 1.75 per cent. 189 Exhibit 22570-X0767.01, PDF page 29. 190 Decision 20622-D01-2016, paragraph 51. 191 Exhibit 22570-X0878.

0

0.5

1

1.5

2

2.5

3

3.5

1 month 3 month 6 month 1 year 5 year 10 year 30 year

Canadian

U.S.



157. Dr. Cleary observed that U.S. rates exceeded Canadian rates across the entire yield curve.

At the long end of the yield curve, U.S. rates exceeded those in Canada by approximately

66 basis points (bps) for 10-year bonds and 78 bps for 30-year bonds.192 Dr. Cleary further

explained that according to the 10-year government yield forecasts for Canada and the U.S. from

the Consensus Economics forecasts in October 2017, the spread between U.S. and Canadian

rates is expected to narrow “to 40 bps by October of 2018.”193

158. Mr. Coyne observed that while the yields for 10-year and 30-year GOC bonds increased

from January 2016 to August 2017, the spreads between these 10-year and 30-year GOC bonds

have decreased from 79 bps in January 2016 to 43 bps in August 2017, which is below the

historical average of 48 bps from 2002 to 2017.194 Mr. Thygesen referenced several articles

indicating that the U.S. yield curve is flattening with the difference between short-term and long-

term yields being at its lowest since November 2007.195 As of March 16, 2018, the spread

between the 10-year and 30-year GOC bonds was 11 bps, whereas the average for the month of

March 2018 was 17 bps. This information is set out in Figure 2.

Figure 2 Canadian government bond yields, 10-year vs. 30-year196

159. Mr. Buttke,197 Dr. Villadsen,198 Mr. Coyne,199 Mr. Hevert200 and Dr. Cleary201 all agreed

that the current GCOC proceeding took place during a rising interest rate environment.

192 Exhibit 22570-X0562.01, PDF page 23. Exhibit 22570-X0878. 193 Exhibit 22570-X0562.01, PDF page 23. 194 Exhibit 22570-X0131, PDF page 21. 195 Exhibit 22570-X0551, PDF pages 20-25. 196 Exhibit 22570-X0835. 197 Exhibit 22570-X0179, PDF page 59. Transcript, Volume 3, pages 574-575. 198 Exhibit 22570-X0193.01, PDF page 23. 199 Exhibit 22570-X0131, PDF page 28. Transcript, Volume 5, page 933. 200 Exhibit 22570-X0153.01, PDF page 10. Transcript, Volume 6, pages 1159-1160. 201 Exhibit 22570-X0562.01, PDF page 22. Transcript, Volume 10, pages 2076-2077.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

20

02

/9

20

03

/2

20

03

/7

20

03

/12

20

04

/5

20

04

/10

20

05

/3

20

05

/8

20

06

/1

20

06

/6

20

06

/11

20

07

/4

20

07

/9

20

08

/2

20

08

/7

20

08

/12

20

09

/5

20

09

/10

20

10

/3

20

10

/8

20

11

/1

20

11

/6

20

11

/11

20

12

/4

20

12

/9

20

13

/2

20

13

/7

20

13

/12

20

14

/5

20

14

/10

20

15

/3

20

15

/8

20

16

/1

20

16

/6

20

16

/11

20

17

/4

20

17

/9

20

18

/2

YIE

LDS (

%)

Canadian Government 10-Year Note Canadian Government 30-Year Bond

Average Historical Spread (Sep 2002 - Mar 2018) = 0.47%

Aug 2017 Average Spread = 0.43% Mar 2018 Average Spread = 0.17%



Looking forward, these witnesses all agreed that 10-year and 30-year GOC bond yields are

expected to increase; however, they disagreed on the timing and magnitude of the expected

increases over the test period.

160. These same witnesses also agreed that central banks raising their policy interest rates

together with increasing inflation expectations are causing short-term interest rates to rise, but

that these are only some of the factors. Dr. Cleary indicated that the Bank of Canada is expected

to raise its policy interest rate one or two more times in 2018.202 However, Dr. Cleary pointed out

that just because the U.S. 10-year yields go up does not necessarily mean the GOC 30-year

yields will go up,203 and that at the time of this proceeding the GOC 30-year bond yields have

remained low despite increases in short-term interest rates.204 Similarly, the CCA indicated that

only the short-term rates are increasing, adding that long-term rates are lower than they were a

year ago when short-term rates started to increase.205 Mr. Thygesen pointed out that the interest

rates for 10-year U.S. and GOC bonds have overall been on a downward trend since 1990.206

Rising short-term rates and falling long-term rates result in a flattening yield curve.

161. Another cause for the expected increase in short-term interest rates mentioned by the

witnesses was the unwinding of quantitative easing policies in the U.S. and Europe. Mr. Buttke

mentioned that due to the unwinding of U.S. monetary stimulus, the market expects incremental

upward pressure on U.S. Treasury 10-year yields of approximately 40 bps or more during the

2018-2020 GCOC period.207 Dr. Cleary stated that he had no reason to disagree with this

assessment,208 while Mr. Coyne commented that this was a conservative estimate and that the

upward pressure on U.S. Treasury 10-year yields could be as high as 100 bps.209

162. Another cause suggested for the increase in short-term interest rates was increasing

economic growth in North America and globally,210 as this shifts the supply and demand for

money in capital markets.211

6.4 Credit spreads

163. In past GCOC decisions, the Commission has accepted that credit spreads are an

objective measure, based on observable market data, which help inform the Commission about

investors’ risk perceptions.212 In this proceeding, the parties pointed out that credit spreads for the

Canadian A-rated utilities have narrowed since the 2016 GCOC proceeding.

164. Mr. Coyne explained that credit spreads are a measure of the difference between the

yields of different securities, and these are typically expressed as a spread between bonds of the

same maturity, but different quality in terms of risk.213 “Credit spread,” as referred to in this

202 Transcript, Volume 10, page 2077. 203 Transcript, Volume 10, page 2080. 204 Transcript, Volume 10, page 2081. 205 Exhibit 22570-X0888, paragraph 107. 206 Exhibit 22570-X0551, PDF page 28. 207 Exhibit 22570-X0179, PDF page 66. 208 Transcript, Volume 10, pages 2079-2080. 209 Transcript, Volume 5, page 911. 210 Transcript, Volume 6, page 1160. 211 Transcript, Volume 5, page 910. 212 Decision 20622-D01-2016, paragraphs 86 and 334. 213 Exhibit 22570-X0131, PDF page 22.



decision, is the difference between the yield on 30-year Canadian A-rated utility bonds and the

yield on 30-year GOC bonds.

165. In the 2016 GCOC decision, the Commission concluded that:

The average credit spread prior to the financial crisis (2001-2007) was around 100 bps,

and the average credit spread after the financial crisis (late 2009-early 2015) remained

relatively stable in the 130 to 150 bps range. In late June 2015, credit spreads began to

widen above 150 bps and reached 190 bps by the end of 2015. Credit spreads then

increased further to 206 bps by February 3, 2016, before declining to about 170 bps as of

the start of the oral hearing in late May 2016. Thus, Dr. Villadsen, Dr. Booth and

Dr. Cleary pointed out that, at the start of the current proceeding, the credit spread was

elevated by some 100 bps relative to what they considered to be its typical or “normal”

level. Mr. Hevert pointed out that credit spread volatility has increased as well.

166. Specifically, as demonstrated in figures 3 and 4 below, the credit spread was 179 bps at

the close of record of the 2016 GCOC proceeding versus 130 bps in March 2018, at the time of

the hearing for this proceeding, a decrease of 49 bps.

167. Dr. Cleary and Dr. Villadsen agreed that the credit spread still remained slightly elevated

compared to the typical or normal level prior to the financial crisis, although both expected it

may continue to narrow.214 Mr. Coyne and Mr. Hevert were of the view that the credit spread is

currently at normal levels.215 Mr. Coyne considered that the credit spread will remain stable over

the test period,216 while Mr. Hevert expressed the view that the credit spread may widen.217

Mr. Thygesen concluded that even if interest rates are to rise, the effect on the credit spread is

expected to be muted and can be expected to mitigate the impact of rising rates on utilities.218

214 Transcript, Volume 3, pages 577-579. Transcript, Volume 10, page 2086. 215 Transcript, Volume 5, pages 915-916. Transcript, Volume 6, page 1165. 216 Transcript, Volume 5, page 919. 217 Transcript, Volume 6, pages 1166-1167. 218 Exhibit 22570-X0551, paragraph 51.



Figure 3 30-year Canadian A-rated utility bond yields, 30-year GOC bond yields219

Figure 4 Credit spread between 30-year Canadian A-rated utility bond yields and 30-year GOC bond yields220

219 Exhibit 22570-X0835. Exhibit 22570-X0836. 220 Exhibit 22570-X0835.

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

20

02

/9

20

03

/2

20

03

/7

20

03

/12

20

04

/5

20

04

/10

20

05

/3

20

05

/8

20

06

/1

20

06

/6

20

06

/11

20

07

/4

20

07

/9

20

08

/2

20

08

/7

20

08

/12

20

09

/5

20

09

/10

20

10

/3

20

10

/8

20

11

/1

20

11

/6

20

11

/11

20

12

/4

20

12

/9

20

13

/2

20

13

/7

20

13

/12

20

14

/5

20

14

/10

20

15

/3

20

15

/8

20

16

/1

20

16

/6

20

16

/11

20

17

/4

20

17

/9

20

18

/2

YIE

LDS (

%)

Canadian Corporate Utility A Index Canadian Government 30-Year Bond

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

20

02

/9

20

03

/2

20

03

/7

20

03

/12

20

04

/5

20

04

/10

20

05

/3

20

05

/8

20

06

/1

20

06

/6

20

06

/11

20

07

/4

20

07

/9

20

08

/2

20

08

/7

20

08

/12

20

09

/5

20

09

/10

20

10

/3

20

10

/8

20

11

/1

20

11

/6

20

11

/11

20

12

/4

20

12

/9

20

13

/2

20

13

/7

20

13

/12

20

14

/5

20

14

/10

20

15

/3

20

15

/8

20

16

/1

20

16

/6

20

16

/11

20

17

/4

20

17

/9

20

18

/2

YIE

LDS (

%)

Spread Average Historical Spread (Sep 2002 - Mar 2018)

2/28/2015 6/30/2016 8/31/2017 3/21/2018

CAN 30Y Govt. Bond 1.92 1.72 2.26 2.37 CAN Utility (A) 3.35 3.50 3.65 3.67

Spread 1.43 1.79 1.38 1.30

Average Historical Spread (Sep 2002 - Mar 2018) = 1.36%

Aug 2017 Average Spread = 1.35% Mar 2018 Average Spread = 1.30%



168. Mr. Buttke provided data that the credit spread for certain Alberta utilities has narrowed

since the time of the 2016 GCOC proceeding.221 Mr. Hevert222 provided similar information for

other Alberta utilities. This is shown in Figure 5 below.

Figure 5 30-year credit spreads for Alberta utilities223

169. The Commission asked in its final issues list if there would be a “clear and objective

measure on the record by which the Commission can determine which factor or factors explain

any changes in utility credit spreads.”224 To this, Mr. Coyne responded:

Though credit spreads provide information on the overall level of perceived risk in the

market, and changes or trends in credit spreads can be meaningful in assessing investors’

required returns, credit spreads are the product of a variety of complex market influences

impacting both the underlying security (e.g., treasury yield), and the security being

measured (e.g., a 30-year A rated utility bond yield). Spreads typically move higher when

there is greater risk of default in the sector, or in the economy as a whole, and vice versa,

as default risk decreases. But this is not the only factor affecting spreads. Investor

demand for bonds of differing quality and risk in relation to other investment options also

plays a role. For these reasons, credit spreads are a relative indicator, a culmination of

market information as it pertains to government and corporate yields, but cannot be

quantified by a specific set of factors.225

221 Exhibit 22570-X0179, PDF pages 63-64. Exhibit 22570-X0815. 222 Exhibit 22570-X0863. 223 Underlying data provided in exhibits 22750-X0816 and 22750-X0864. 224 Exhibit 22570-X0078, paragraph 3. 225 Exhibit 22570-X0131, PDF page 24.

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

Ap

r-1

3

Jun

-13

Au

g-1

3

Oct

-13

Dec

-13

Feb

-14

Ap

r-1

4

Jun

-14

Au

g-1

4

Oct

-14

Dec

-14

Feb

-15

Ap

r-1

5

Jun

-15

Au

g-1

5

Oct

-15

Dec

-15

Feb

-16

Ap

r-1

6

Jun

-16

Au

g-1

6

Oct

-16

Dec

-16

Feb

-17

Ap

r-1

7

Jun

-17

Au

g-1

7

Oct

-17

Dec

-17

Feb

-18

Altalink EPCOR Utilities FortisAB CU Inc. AltaGas



170. Dr. Cleary stated that changes in government yields and yield spreads tend to go in

opposite directions, and offset one another to a certain extent.226 In addition, he calculated that

the correlation coefficient between 30-year GOC bonds and A-rated utility yield spreads over the

January 2003 to November 2017 period was -0.49, which indicates a strong negative

relationship.227 Mr. Hevert commented that while credit spreads and interest rates are inversely

related over longer horizons, within shorter periods that relationship may be less stable.228

Mr. Buttke expressed a similar view.229

171. In the 2016 GCOC proceeding, Mr. Hevert concluded that there was little question that

the increase in credit spreads suggested some measure of increased risk perception among

Canadian utility investors.230 However, AltaLink, EPCOR and Fortis, relying on Mr. Hevert’s

evidence in this proceeding, came to a different conclusion in the current proceeding. They

submitted that although credit spreads have narrowed since the 2016 GCOC proceeding, that is

not a basis for concluding that the risk perceptions of utility equity investors have decreased.231

172. In contrast, the UCA pointed out that:

… in the 2016 GCOC proceeding, the utilities’ witnesses focused on elevated utility

credit spreads, while ignoring the impact of prevailing low interest rates. In this

proceeding, the utilities’ witnesses now heavily stress the anticipated (but far from

certain) rise in interest rates, while ignoring or heavily downplaying the significance of

the notable decrease in utility credit spreads. This is so notwithstanding the net impact in

both scenarios is similar – i.e. low borrowing costs for utilities.232

173. Given that debt financing for Alberta’s utilities remains at historic lows, the UCA

concluded that the cost of equity must also be similarly low, on a relative and absolute basis,

given the strong relationship between the cost of debt and the cost of equity.233

6.5 Market volatility

174. In the 2016 GCOC proceeding, Mr. Hevert, Dr. Villadsen, Dr. Cleary and Dr. Booth

drew the Commission’s attention to the fact that stock market volatility had increased in late

2015 and early 2016.234 In particular, two measures of the market’s expectations for volatility

were relied upon during that proceeding to demonstrate this point: (1) the VIXC, which measures

the 30-day implied volatility of the Standard & Poor’s (S&P) Toronto Stock Exchange (TSX)

60 index (representing the stock market in Canada); and (2) the VIX, which measures the 30-day

implied volatility of the S&P 500 index (representing the stock market in the U.S.). During the

2016 GCOC proceeding, witnesses explained to the Commission that these indexes are “highly

226 Exhibit 22570-X0562.01, PDF page 27. 227 Exhibit 22570-X0562.01, PDF page 14. 228 Exhibit 22570-X0741.01, PDF page 13. 229 Exhibit 22570-X0749, PDF page 91. 230 Decision 20622-D01-2016, paragraph 64. 231 Exhibit 22570-X0890.01, paragraph 29. 232 Exhibit 22570-X0913, paragraph 11. 233 Exhibit 22570-X0897.01, paragraph 36. 234 Decision 20622-D01-2016, paragraph 68.



visible, and often-reported barometers of investor risk sentiments” and are often referred to as

the “investor fear gauge.”235

175. In the 2016 GCOC proceeding, the witnesses agreed that the long-term average for both

the VIXC and VIX was about 20.236 They further pointed out that volatility stayed at relatively

low levels during 2013 and 2014, but in August 2015, the VIXC and VIX spiked to 33 and 40,

levels not seen since October 2011, and in January 2016 volatility remained elevated and stood at

about 26 for both indices.237 At the close of record for the 2016 GCOC proceeding, the VIXC and

VIX were approximately 13 and 16, respectively, as shown in Figure 6.

176. In the 2016 GCOC proceeding, the Commission concluded that the observed instability

in estimators of investor perceptions of near-term market uncertainty, like the VIX and the

VIXC, were indicative of increased investor uncertainty in the 2016-2017 period compared to

investor uncertainty at the time of the 2013 GCOC proceeding.238

177. In the current proceeding, Mr. Coyne,239 Mr. Buttke,240 Dr. Villadsen,241 Mr. Hevert242 and

Dr. Cleary243 all provided evidence on the VIXC and VIX. As shown in Figure 6, the VIXC and

the VIX stood at approximately 10 and 11, respectively, at the end of August 2017. These levels

spiked briefly in early February 2018, reaching levels of approximately 22 and 33, respectively.

At the end of March 2018, during the time of the hearing, these levels were roughly 12 and 18,

respectively.244

235 Decision 20622-D01-2016, paragraph 68. 236 Decision 20622-D01-2016, paragraph 69. 237 Decision 20622-D01-2016, paragraph 69. 238 Decision 20622-D01-2016, paragraph 91. 239 Exhibit 22570-X0131, PDF pages 24-25. 240 Exhibit 22570-X0179, PDF pages 47-49. 241 Exhibit 22570-X0193.01, PDF pages 28-29. 242 Exhibit 22570-X0153.01, PDF pages 31-33. 243 Exhibit 22570-X0562.01, PDF page 16. 244 Transcript, Volume 10, page 2097.



Figure 6 Canadian and U.S. stock market volatility indexes245

178. While the VIXC and VIX have generally decreased since January 2016, Mr. Coyne,246

Dr. Villadsen247 and Mr. Hevert248 did not agree that this was indicative of lower volatility in the

market. All of these witnesses reminded the Commission that the VIXC and VIX are “near-term”

measures of market volatility, extending out 30 days, and they each pointed to alternative

indicators of volatility, such as the State Street Investor Confidence Indices,249 the SKEW

Index,250 and the term structure of volatility of the Chicago Board Options Exchange.251 During

the hearing, Mr. Buttke confirmed that he did not provide evidence on the SKEW Index during

the 2016 GCOC proceeding.252

179. Dr. Cleary underscored that there is always a certain amount of volatility in the market

and suggested that rather than focus on temporary spikes, the VIX and VIXC should be

examined over a period of time. He further suggested that since these values have not remained

elevated over a sustained period of time, this may indicate market anxiety is above normal

levels.253 Dr. Cleary also referred to alternative indicators of market risk, including the Mercer

245 Exhibit 22570-X0817. 246 Transcript, Volume 5, pages 923-926. 247 Transcript, Volume 3, page 586. 248 Transcript, Volume 6, page 1176. 249 Exhibit 22570-X0131, PDF page 26. 250 Exhibit 22570-X0193.01, PDF page 30. Exhibit 22570-X0179, PDF pages 49-51. 251 Exhibit 22570-X0153.01, PDF pages 34-35. 252 Transcript, Volume 3, page 588. 253 Transcript, Volume 10, page 2098.



Pension Health Index,254 the trailing price-earnings ratio for the S&P/TSX Composite Index, the

U.S. S&P 500 Index255 and the Financial Stress Index.256

180. Mr. Thygesen stated:

The VIX and VIXC are basically half the levels of 2016. This has led to shift in utility

evidence basically saying, ‘yes but look what is lurking around the corner’. In my view

the treatment of the evidence should be consistent. If the weight was on current

conditions in 2016 then the weight should be on current conditions now.257

181. Mr. Thygesen pointed to yet other indicators that show lower volatility in the market,

including the Chicago Fed National Financial Conditions Index, the Kansas City Financial Stress

Index and the St. Louis Financial Stress Index.258

182. Interveners argued that the VIXC and VIX are at generally lower levels than during the

2016 GCOC proceeding, and that gives weight to a decrease in the approved ROE.259 The utilities

pointed to increases in the VIXC and VIX observed in February 2018, stating that because of the

movement in these indices, the argument that the approved ROE should be lowered because of

lower market volatility is no longer defensible.260

183. During the hearing, Dr. Cleary agreed with recent statements made by a Bank of Canada

deputy governor that more normal levels of volatility are returning to markets. Dr. Cleary

elaborated, saying “The VIX [and VIXC] were a little bit below average on Tuesday [March 20,

2018], and they were well below average in the fall. So it seems to be the case, we have

volatility. And there's always going to be volatility in the market.”261

6.6 Overall conclusions of the witnesses

184. Mr. Buttke’s view was that global markets have been strong and are likely to continue to

strengthen in the future. He pointed to central banks raising policy rates, with additional rate

hikes predicted. Quantitative easing programs are being reversed in the U.S. and Europe, which

will further cause interest rates to rise.262

185. Dr. Villadsen observed that both utility bond yields and government bond yields are

expected to increase over the next several years. She pointed to a spike in the VIXC and VIX in

February 2018, reminding the Commission that they are just one measure and that “they are one-

month-ahead indicators of volatility, and they’re meaningful in that sense. They’re not

meaningful in a long-term sense.”263

254 Exhibit 22570-X0562.01, PDF page 16. 255 Exhibit 22570-X0562.01, PDF pages 15-16. 256 Transcript, Volume 10, page 2098. 257 Exhibit 22570-X0551, PDF page 5. 258 Exhibit 22570-X0551, PDF pages 49-51. 259 Exhibit 22570-X0897.01, PDF pages 16-17. Exhibit 22570-X0888, PDF page 20. 260 Exhibit 22570-X890.01, PDF pages 18-19. Exhibit 22570-0900, PDF pages 24-25. 261 Transcript, Volume 10, pages 2097-2098. 262 Exhibit 22570-X0179, PDF pages 4-6. 263 Transcript, Volume 3, page 164.



186. Mr. Hevert summarized his view by stating that observed and expected interest rates have

increased and economic growth has improved. He stated that these factors, taken in conjunction

with his view that business risks have not diminished, support his recommendation for an

increased approved ROE.264

187. Mr. Coyne considered global economic growth to be about the same as in the 2016

GCOC proceeding and on a stable trend.265 The Fed’s and the Bank of Canada’s key interest rates

are on an upward trend, and the yields on long-term government bonds have increased since

2016 and are expected to increase further.266

188. Dr. Cleary summarized his views as follows:

Both Canada and Alberta are expected to experience more moderate but solid GDP

growth going forward. Bond yield spreads have declined, as has stock market volatility,

and both bond and stock markets are healthy. In other words, economic and capital

market conditions are solid today, improved since 2016, and far removed from those

existing at the peak of the 2008-2009 financial crisis.267

189. Mr. Thygesen’s view was that virtually all risk measures are lower than they were in the

2016 GCOC proceeding, including the VIX and VIXC, which are basically half the levels of

2016. Further, his view was that utility spreads have decreased substantially since the period,

which is consistent with the lower risk measures.268

6.7 Commission findings

190. In Decision 20622-D01-2016, the Commission found that the global and Canadian

economic capital market conditions present at that time were different from the conditions that

existed during the global financial crisis of 2008-2009, but there were still lingering effects of the

global financial crisis.269 Further, the Commission found that economic conditions were generally

expected to improve in 2017, including an expected increase in interest rates and utility bond

yields. The Commission also recognized that credit spreads had widened and market volatility

was elevated compared to the 2013 GCOC proceeding.270 Given all of the evidence, the

Commission found that an increase in approved ROE was warranted for 2017.271

191. In the current proceeding, the Commission observes that Canadian actual real GDP for

2016 and 2017 was 1.4 and 3.0 per cent, respectively;272 inflation and interest rates have risen in

2017, while utility bond yields remain effectively unchanged since the 2016 GCOC proceeding.

192. Based on this and other evidence filed on this proceeding, the Commission finds that the

global economic and Canadian capital market conditions have improved since the time of the


264 Exhibit 22570-X0153.01, PDF page 10. 265 Transcript, Volume 5, page 905. 266 Transcript, Volume 5, page 909. 267 Exhibit 22570-X0562.01, PDF page 5. 268 Exhibit 22570-X0551, PDF page 5. 269 Decision 20622-D01-2016, paragraph 81. 270 Decision 20622-D01-2016, paragraphs 86, 89-90, 150-151. 271 Decision 20622-D01-2016, paragraph 337. 272 Exhibit 22570-X0749, PDF page 104.



193. A recent speech by a Bank of Canada deputy governor filed on the record by Mr. Buttke

provides a succinct summarization of the current global economic and Canadian capital market

conditions: “Canada is a very open economy, and its growth is supported by what is now a

synchronous global expansion. We are now seeing solid growth not only in the United States and

China but also in Europe, as well as in many other emerging-market economies.”273 Therefore,

the Commission agrees with Dr. Cleary who expressed the view that economic and capital

market conditions are “far removed from those existing at the peak of the 2008-2009 financial

crisis.”274

194. Looking forward, the Commission was presented with forecasts of Canadian economic

growth, including projections by the Bank of Canada, that indicate slowing economic growth,

with rates of 2.1 per cent in 2018 and 1.5 per cent in 2019.275 Inflation is broadly expected to be

near the Bank of Canada’s target rate of two per cent over this same period.276

195. However, the Commission recognizes that future growth expectations are far from certain

and are dependent on many factors, both domestic and international. Stated in reference to

market volatility, ATCO and AltaGas argued that “the Commission should give weight to the

Bank’s [Bank of Canada] longer-term view of increased volatility expectations.”277

196. A strong example of market uncertainty present during this proceeding that may cause

both short-term and long-term volatility is the uncertain outcome of NAFTA negotiations. On

this, the Commission finds the Bank of Canada’s view, useful:

… uncertainty about the North American Free Trade Agreement (NAFTA) and growing

global trade tensions will need to be watched, for their possible impact on the outlook.

Recent developments with respect to steel and aluminum, alongside increased

protectionist rhetoric, carry potentially serious consequences. We do not know how or

when the NAFTA talks or other trade disputes will conclude, and we do not know how

industries, or governments, will react. The range of possibilities is wide, which means

that trying to quantify any scenario in advance would not be useful for monetary policy

purposes. For now, our working assumption is that existing trade arrangements will stay

in place over our current two-year projection horizon. As and when concrete outcomes

emerge, we will be in a better position to assess their impact on the Canadian economy.278

197. The Commission shares the broadly held view that the current proceeding was during a

period of rising short-term interest rates.279 It is readily observable that the Bank of Canada and

the Fed were raising their policy interest rates, with the Bank of Canada having done so three

times and the Fed having done so five times between the 2016 GCOC proceeding and the close

273 Exhibit 22570-X0823. 274 Exhibit 22570-X0562.01, PDF page 5. 275 Exhibit 22570-X0562.01, PDF page 19. 276 Exhibit 22570-X0562.01, PDF page 20. Exhibit 22570-X0918, paragraphs 33 and 54. Exhibit 22570-X0909,

paragraph 79. 277 Exhibit 22570-X0900, paragraph 25. 278 Exhibit 22570-X0823, PDF page 6. 279 Transcript, Volume 3, page 575. Transcript, Volume 5, page 909. Transcript, Volume 6, page 1159. Transcript,

Volume 10, pages 2071-2080.



of record for this proceeding. At the close of record for this proceeding, additional rate increases

were expected in the remainder of 2018.280

198. What is not as readily transparent is how these rising short-term interest rates should

impact the Commission’s determinations in setting the 2018 to 2020 ROE. As Dr. Cleary pointed

out, central bank policy interest rates only tend to affect the “short end” of the yield curve281

(Figure 1). The Commission observes that the yield curve is “flattening,” as in, the “long end” of

the yield curve, or the yield on 30-year GOC bonds has not increased to the same degree as

short-term interest rates since the 2016 GCOC proceeding (figures 1 and 2).

199. Using information filed on the record of this proceeding, the Commission has plotted the

recent movements of the 10-year and 30-year GOC bonds and the 30-year utility bond yields in

Figure 7 and notes the following:

The 30-year GOC bond yields have not increased to the same extent as the 10-year GOC

bond yields and the spread between them has contracted to 17 bps at the time of the

hearing for this proceeding, compared to the long-run historical average of approximately

50 bps.

While the 30-year GOC bond yields have increased slightly since the close of record for

the 2016 GCOC proceeding, when considering their movement over the last three years,

they are generally unchanged.

The 30-year utility bond yields have stayed in the range of those present during the 2016

GCOC proceeding. This has had the effect that credit spreads between 30-year utility

bond yields and 30-year GOC bond yields have decreased from 179 bps at the time of the

hearing for the 2016 GCOC proceeding to 130 bps at the time of the hearing for this

proceeding.

280 Transcript, Volume 3, page 561. Transcript, Volume 5, page 909. Transcript, Volume 6, page 1155. Transcript,

Volume 10, pages 2071-2072. 281 Transcript, Volume 6, page 1160.



Figure 7 10-year, 30-year GOC bonds and the 30-year utility bond yields282

200. During the current proceeding, the Commission was presented with different views with

respect to what constitutes the “normal” credit spread and witnesses’ expectations on how the

credit spread will change directionally over this GCOC term. Without engaging in the debate as

to what constitutes the normal level for credit spreads (100 bps according to Dr. Cleary and

Dr. Villadsen, or 130 bps according to Mr. Coyne and Mr. Hevert), the Commission continues to

hold the view that credit spreads at the time of the 2016 GCOC proceeding were elevated

compared to any level considered normal. Since the 2016 GCOC proceeding, credit spreads have

narrowed significantly.

201. The Commission continues to be of the view that credit spreads are an objective measure,

based on observable market data, which help to inform the Commission about utility bond

investors’ risk perceptions, and by implication, to some extent, the expectations of utility equity

investors. While evidence was put forward by Mr. Hevert that a decline in credit spreads may be

of a short-term, temporary nature, and may not be indicative of a change in risk perceptions in

the market,283 the Commission is not persuaded by this evidence, which is contradicted by his

own claims during the 2016 GCOC proceeding that the increase in credit spreads at that time

demonstrated an increase in investors’ risk perceptions.284

282 Underlying data taken from Exhibit 22570-X0836. 283 Exhibit 22570-X0741, PDF pages 14-18. 284 Decision 20622-D01-2016, paragraph 64.

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.02

01

5/1

20

15

/2

20

15

/3

20

15

/4

20

15

/5

20

15

/6

20

15

/7

20

15

/8

20

15

/9

20

15

/10

20

15

/11

20

15

/12

20

16

/1

20

16

/2

20

16

/3

20

16

/4

20

16

/5

20

16

/6

20

16

/7

20

16

/8

20

16

/9

20

16

/10

20

16

/11

20

16

/12

20

17

/1

20

17

/2

20

17

/3

20

17

/4

20

17

/5

20

17

/6

20

17

/7

20

17

/8

20

17

/9

20

17

/10

20

17

/11

20

17

/12

20

18

/1

20

18

/2

20

18

/3

YIE

LDS (

%)

Canadian Corporate Utility A Index Canadian Government 30-Year Bond

Canadian Government 10-Year Note

2016 GCOC close of record: 30-yr utility 3.4230-yr GOC 1.67Spread (utility-30 yr) 1.75

10-yr GOC 1.05Spread (30-10 yr) 0.61

2018 GCOC hearing: 30-yr utility 3.6730-yr GOC 2.37Spread (utility-30 yr) 1.30

10-yr GOC 2.02Spread (30-10 yr) 0.17

Start of oral hearing for Proceeing 20622

Release of Decision 20622-D01-2016



202. As further discussed in Section 8.3 of this decision, Dr. Cleary,285 Mr. Hevert and

Mr. Coyne presented evidence showing a negative correlation between changes in government

yields and yield spreads, and that they tend to offset one another to a certain extent. With respect

to future interest rates, witnesses provided evidence that 30-year GOC bond yields are generally

expected to increase;286 however, the Commission shares Dr. Cleary’s view that “this is far from

a given fact,”287 as exhibited by the “flattening” yield curve (Figure 1). In contrast, witnesses

filed less evidence on future expectations for 30-year utility bonds, which were near five-year

lows (Figure 3). Given all the variables that may affect credits spreads and the disparate views

held by witnesses on future expectations, the Commission is not able to arrive at a conclusion

regarding how credit spreads will move directionally over the 2018-2020 period, other than they

will likely militate any move in the underlying 30-year GOC bond yield.

203. In the 2016 GCOC proceeding, the Commission was presented with evidence that

estimators of investor perceptions of near-term market uncertainty, particularly the VIX and the

VIXC, were indicative of increased investor uncertainty in the 2016-2017 period compared to

investor uncertainty which existed at the time of the 2013 GCOC proceeding.288 The VIX and

VIXC were presented during the current proceeding, along with additional estimators of investor

perceptions that were not before the Commission in the 2016 GCOC proceeding.

204. The Commission accepted the VIX and VIXC as estimators of investor perceptions of

volatility and gave them weight in determining the level of market volatility in the 2016 GCOC

proceeding,289 and no party has satisfied the Commission that anything has changed since the

2016 GCOC proceeding to depart from this.

205. The Commission observes that, based on Figure 6, the VIX and VIXC were relatively

stable at or below longer-term averages since the 2016 GCOC proceeding, apart from a

temporary spike in February 2018. The Commission is of the view that while some amount of

volatility will always exist in the market, the level of volatility is lower than at the time of the

2016 GCOC proceeding. However, given that the VIX and VIXC are short-term measures of

volatility, they do not necessarily provide an indication of investor uncertainty for 2019 and

2020.

206. In conclusion, the Commission finds that the global economic and Canadian capital

market conditions have improved since the 2016 GCOC proceeding, and are far removed from

the 2008-2009 financial crisis. In particular, the Commission observes that there has been global

and national economic growth, reduced market volatility, a modest increase in the 30-year GOC

bond yield and a compression in credit spreads. However, as will be discussed in the sections

that follow, the Commission finds that the upward pressure associated with certain of these

factors is largely offset by the downward pressure associated with others. On balance, these

285 Exhibit 22570-X0562.01, PDF page 27. 286 Exhibit 22570-X0179, PDF page 59. Transcript, Volume 3, pages 574-575. Exhibit 22570-X0193.01, PDF

page 23. Exhibit 22570-X0131, PDF page 28. Transcript, Volume 5, page 933. Exhibit 22570-X0153.01,

PDF page 10. Transcript, Volume 6, pages 1159-1160. Exhibit 22570-X0562.01, PDF page 22. Transcript,

Volume 10, pages 2076-2077. 287 Exhibit 22570-X0562.01, PDF page 25. 288 Decision 20622-D01-2016, paragraphs 90-91. 289 Decision 20622-D01-2016, paragraph 91.



factors indicate the approved ROE for 2018 should be at or near that set in the 2016 GCOC

decision.

207. The Commission has also considered the evidence filed on the record of this proceeding

with respect to future expectations for global economic and Canadian capital market conditions.

Given the expectations of diminishing national GDP growth rates, moderately higher inflation to

reach the mid-point of the Bank of Canada’s target range, increasing short-term interest rates, a

flattening yield curve, but uncertain long-term interest rates and market uncertainty with respect

to international trade, the Commission finds that these factors result in a similar offset and

together indicate that the approved ROE for 2019 and 2020 should be the same or similar to the

value set for 2018.

7 Municipally owned utilities

208. In its July 5, 2017 correspondence, the Commission indicated that it intended to explore

a number of issues in relation to municipally owned utilities in this proceeding:

36. … the Commission considers that the 2018 GCOC proceeding is also a forum to

consider matters with respect to the municipally owned utilities, specifically. The

Commission wishes to explore how their ownership structure and the relationship

between the utilities’ ratepayers and the municipality’s taxpayers may affect ROE and

deemed equity ratios for these utilities. In this regard, the Commission invites

submissions from parties regarding what municipal ownership entails with regard to debt

availability through the Alberta Capital Financing Authority (ACFA), credit metrics in

light of available debt through ACFA, income tax status, the opportunity for municipal

riders and the effect of these factors on the risk profile of the municipally owned

utilities.290

209. In this section, the Commission will address generally the interplay between ownership

structure and the stand-alone principle as it relates to municipally owned utilities, more

specifically the role of ACFA funding in assessing the credit metrics of ENMAX and EPCOR

given the stand-alone principle, and finally the use of equity funding riders. The Commission has

addressed the issue of a utility’s taxable status in relation to capital structure in Section 8.

7.1 Ownership structure and stand-alone principle

210. Both EPCOR and ENMAX are municipally owned corporations. EPCOR’s ultimate

owner, through its parent EPCOR Utilities Inc. (EUI), is the City of Edmonton.291 ENMAX is

wholly owned by ENMAX Corporation which, in turn, is wholly owned by Calgary.292 Both

EPCOR and ENMAX provided evidence and argument on the issues identified by the

Commission, as noted above. While Red Deer and Lethbridge also operate municipal electric

utilities, neither operate the utility as a separate corporate entity nor did either raise

considerations with respect to the stand-alone principle in this proceeding. ENMAX submitted

that the ownership structure of a municipal utility, and the relationship between the utility’s

ratepayers and the municipality’s taxpayers, is not a relevant consideration in determining the

290 Exhibit 22570-X0114. 291 Exhibit 22570-X0195, paragraph 4. 292 Exhibit 22570-X0129, paragraph 4.



cost of capital of a municipally owned utility. It stated that municipally owned utilities must be

regulated on a stand-alone basis.293

211. Mr. Coyne referred to the stand-alone principle for guidance on this issue. He submitted

the stand-alone principle dictates that it is the use for the capital that a cost is applied to, and not

the source.294

212. EPCOR stated that Standard & Poor’s (S&P) and DBRS Limited (DBRS) have each

addressed EUI’s municipal ownership in recent credit-rating reports. DBRS commented that

EUI’s ownership structure limits its ability to access equity markets directly. S&P commented

that there is a low likelihood that the City of Edmonton would provide timely and sufficient

extraordinary support in the event that EUI faces financial distress. S&P added that if EUI

required long-term support in a financial stress scenario, its belief is that EUI would more likely

be sold than receive taxpayer support.295

213. EPCOR added that EUI was originally established in 1996 as a fully independent, stand-

alone subsidiary corporation, and that the City of Edmonton limits its activities in relation to EUI

to that of a shareholder. EUI is governed and managed independently of the City of Edmonton.296

214. EPCOR submitted that the stand-alone principle is fundamental in utility regulation, and

requires that, regardless of who the owner of a utility happens to be, the ROE for that utility must

be established based on what is necessary to attract investment in that utility, having regard for

the risk of the utility on a stand-alone basis.297 ENMAX expressed a similar view, and submitted

that failure to apply the stand-alone principle can result in improper cross-subsidization.298

215. The UCA indicated that the City of Edmonton appoints EPCOR’s board of directors and

external auditors, approves the dividends the City of Edmonton receives from EUI through the

dividend policy, and approves any material asset dispositions. It stated that the City of Edmonton

is undoubtedly involved in the business operations of EPCOR.299

216. In argument, the CCA noted that both Mr. Madsen and Mr. Thygesen agree that the fair

return should reflect the risk of the business itself and not the source of the financing.300

Mr. Madsen also expressed the view, however, that the stand-alone principle should not be

applied “by rote” if a decision causes harm to ratepayers or otherwise is not in the public

interest.301

Commission findings

217. The stand-alone principle has been applied by the Commission to treat a regulated utility

as a distinct entity for the purposes of determining the costs to be borne by ratepayers for the

service of the regulated utility. As noted by the Alberta Court of Appeal in ATCO Electric Ltd.

293 Exhibit 22570-X0896, paragraph 142. 294 Transcript, Volume 5, page 997. 295 Exhibit 22570-X0195, paragraphs 55-57. 296 Exhibit 22570-X0195, paragraphs 36-38. 297 Exhibit 22570-X0195, paragraph 39. 298 Exhibit 22570-X0896, paragraphs 20-22. 299 Exhibit 22570-X0767.01, paragraph 330. 300 Exhibit 22570-X0888, paragraph 384. 301 Exhibit 22570-X0888, paragraph 382.



v Alberta (Energy and Utilities Board), “The purpose of the stand-alone principle is to notionally

isolate and categorize – for accounting and rate-making purposes – the costs incurred in the

operation of a discrete business function of a utility.”302 The principle has been applied to allocate

costs between regulated and non-regulated activities of an integrated utility, with the theory

being that regulated utility customers should only pay for the costs of the regulated service. It has

also been applied to allocate costs incurred by an integrated utility amongst its various business

functions, so that just and reasonable rates can be set for each business function. In the context of

a GCOC proceeding, the stand-alone principle has been applied to determine an ROE and

deemed equity structure for each regulated utility as if it were a stand-alone entity.

218. The stand-alone principle arises with respect to municipal utilities when the Commission

considers what regard, if any, should be had for the fact that the utility is ultimately owned by a

municipality. A municipality possesses certain unique traits that distinguish it from a non-

municipal corporation. For example, a municipality has access to low cost ACFA financing and

the ability to use an equity funding rider, which it may pass on to the regulated utility.

Additionally, a municipally owned utility is exempt from paying income taxes.

219. In some cases, the unique trait(s) of the municipality ultimately flow through to

ratepayers. For example, Calgary makes ACFA financing available to ENMAX, and ratepayers

benefit from ENMAX obtaining financing at lower interest rates than what it could procure

itself.

220. In other cases, the unique features that may be associated with municipal ownership are

not made available to the municipally owned utility and thereby do not flow through to

ratepayers. In contrast to Calgary’s practice for ENMAX, the City of Edmonton does not make

ACFA financing available to EPCOR. Accordingly, in the case of EPCOR, ratepayers pay for

financing at higher interest rates than what would be paid if EPCOR obtained ACFA financing.

In approving EPCOR’s debt financing costs in the past, the Commission has generally been

persuaded to apply the stand-alone principle and considered the cost of debt for each of EPCOR

transmission and EPCOR distribution as if they were distinct corporate entities. This issue is not

without contention, and has been the subject of dispute amongst parties and considerable scrutiny

by the Commission in past tariff proceedings.

221. In considering the application of the stand-alone principle in this proceeding, the

Commission does not accept the submissions of those parties or witnesses who would have the

Commission rigidly apply this principle. To do so would be inconsistent with past consideration

and application of the stand-alone principle. For example, as noted by the Alberta Court of

Appeal in ATCO Electric Ltd. v Alberta (Energy and Utilities Board):

[178] I also note that the evidence of the Independent Financial Experts to the Board,

Messrs. Demcoe and McCormick (collectively the “IFE”), supports the Board’s

approach. The IFE testified that the stand-alone principle was developed as a shield to

protect customers from higher rates due to subsidization of non-regulated activities.

Therefore, in the IFE’s view, it ought not to be used as a sword to require customers to

pay higher rates simply because of a notional separation of what remained as integrated

business functions. The IFE also argued that the stand-alone principle did not reflect the

reality of how a utility accessed the capital market. When a utility sought financing, this

302 ATCO Electric Ltd. v Alberta (Energy and Utilities Board), 2004 ABCA 215, paragraph 175.



was not done on behalf of some discrete business function in the organization but rather

on behalf of the larger corporate entity itself. For these reasons, the IFE concluded that:

... the Board should “not apply the stand-alone principle by rote. Instead the Board

should deal with the reality, utilize independence of thought, question assumptions

and think through whether an approach that has been applied in the past in different

circumstances should be applied now in new circumstances. Such an approach should

lead the Board to deal with reality and to decline to apply the stand-alone principle to

the detriment of the customers of the [distribution companies]

[179] This is precisely what the Board did. It fully considered a number of separate

issues affecting calculation of carrying costs and examined the business risk elements

inherent in that calculation. Its conclusion was that the business risks, including the

capital recovery risks, associated with the administration of the deferral accounts were,

by their nature, very low: Decision 2001-92 at p.46, AB Vol. II, F127. Further, that risk

was “significantly lower than the business risk of any of the three business functions” of

an integrated utility… Thus, the Board decided that it would be fair and reasonable to

consider the deferral accounts operation as a separate stand-alone business unit but within

the totality of the integrated electric utility as it existed in the year 2000. The Board

recognized that if this were not done, and the deferral accounts operation were treated

purely as a stand-alone business as more than one party had urged at hearing, this would

have “likely led to a windfall for the integrated utility.” The Board also noted, correctly

in my view, that while prudent costs does not mean the lowest possible costs “financing

costs that are unnecessary and inflated, or alternatively, result in windfall profits to the

utility cannot be considered prudent.” These are conclusions which the Board was

entitled to reach on this evidentiary record – and they are conclusions which weigh

heavily in favour of the reasonableness of the Board’s approach.

[180] More fundamentally, though, the question of what financial model to use in

calculating carrying costs of a particular business function of a utility’s operations is

precisely the kind of issue which the Legislature intended to leave to the Board’s

discretion. As noted, an important feature of this analysis is the determination of the level

of business and financial risk associated with a particular function. The fact a utility

chooses to order its affairs in a particular fashion for internal purposes does not immunize

it from Board scrutiny to determine what a fair and appropriate allocation of financing

costs would be for a specific business function regardless of how the utility has structured

its operations.

[181] Nor can a utility complain where the Board recognizes that some aspects of an

integrated utility’s business functions are less risky than others – and calculates financing

costs accordingly. The Board is under no obligation to use an integrated utility’s highest

risk functions as the basis for setting the capital requirements of its lowest risk functions.

That would be to ignore commercial realities. Thus, the Board has the jurisdiction to

segregate business functions of an integrated utility – and determine a notional corporate

organizational model – for purposes of evaluating risk and calculating prudent carrying

costs associated therewith.303

222. While the Commission has generally maintained its practice of determining a deemed

equity ratio for each utility that, when combined with the approved ROE, will achieve target

303 ATCO Electric Ltd. v Alberta (Energy and Utilities Board), 2004 ABCA 215, paragraphs 178-181.



credit ratings in the A-range when assessed on a stand-alone basis, it has tempered this approach

when it has determined, based on the evidence before it, that ignoring the utility’s owner (or

investor) would be inconsistent with other considerations, such as the Commission’s obligation

to ensure rates are just and reasonable. Put another way, while the Commission continues to

apply the stand-alone principle, this is just one tool to assist it in determining a fair return and

approving just and reasonable rates, as detailed in the fair return section above.

223. The Commission identified issues with respect to municipal ownership, such as debt

availability through ACFA and the impact of ACFA on credit metrics, the opportunity for

municipal riders and the effect of these factors on the risk profile of the municipally owned

utilities, as matters to be considered in this proceeding, and the Commission discusses its

findings on these specific issues below. In so doing, the Commission has balanced the

application of the stand-alone principle, as discussed above, with other considerations, including

the fair return standard and the Commission’s overall obligation to ensure that rates are just and

reasonable.

7.2 ACFA funding, credit metrics and stand-alone principle

224. The City of Edmonton and Calgary each have access to funding from ACFA, which is

low cost debt based on the Province of Alberta’s credit rating.304 Calgary makes funding from

ACFA available to ENMAX, and low cost debt from ACFA is passed on to ENMAX

customers.305 The City of Edmonton does not make funding from ACFA available to EPCOR.

225. EPCOR noted that the issue of availability of ACFA financing has arisen before the

Commission or its predecessor on at least three occasions over the last decade. It noted that on

two of these occasions, the Commission directed EPCOR to approach the City of Edmonton and

inquire as to whether Edmonton would make funding from ACFA available to EPCOR. The City

of Edmonton declined to make funding from ACFA available to EPCOR. EPCOR submitted

that, in all cases, the Commission honoured the stand-alone principle and refused to deem

EPCOR’s approved debt rates at the ACFA rates.306 EPCOR stated that the evidence is

uncontroverted that it cannot access ACFA financing.307

226. Mr. Hevert submitted that because EPCOR cannot access ACFA financing, any issues

related to credit metrics and the risk profile arising from the use of such funding are moot.

He stated that even if such financing was available, this would not affect the risk of EPCOR’s

operations, or the return required on its equity.308

227. Mr. Coyne explained that while ENMAX may access funding from ACFA, the

availability of this funding is at the discretion of Calgary. As a result, Mr. Coyne submitted that

304 See, for example, Exhibit 22570-X0195, at paragraph 43, citing Decision 2006-054: EPCOR Transmission Inc.,

2005/2006 Transmission Facility Owner Tariff, EPCOR Distribution Inc., 2005/2006 Distribution Tariff –

Phase I, Applications 1389884-1 and 1389885-1, June 15, 2006. 305 Exhibit 22570-X0129, paragraphs 4, 9. 306 Exhibit 22570-X0195, paragraph 41. 307 Exhibit 22570-X0195, paragraph 52. 308 Exhibit 22570-X0153.01, PDF page 128.



funding from ACFA should not be factored into the cost of equity determination for ENMAX on

a stand-alone basis.309

228. In argument, the CCA did not advocate for asymmetrical application of the stand-alone

principle, submitting:

If a utility avails itself of ACFA funding (ENMAX) and that funding results in improved

credit metrics, the Commission should not award that utility a lesser equity thickness

simply on the basis of the improved credit metrics resulting from the use of ACFA

funding. Similarly, if a utility does not avail itself of ACFA funding (EDTI), and yet that

utility has depressed credit metrics as a result, the Commission should not award that

utility additional equity thickness simply on the basis of the weakened credit metrics

resulting from not using ACFA funding. The CCA does not view this as an asymmetric

application of the stand-alone principles. In both cases the effects of the ACFA funding

are ignored when approving an equity thickness.310

229. Mr. Madsen submitted that EPCOR’s shareholder has decided not to use ACFA funding,

and the result is that EPCOR has higher long-term debt rates and weaker credit metrics. He

submitted that the Commission should consider requiring any future long-term debt issued by

EPCOR to be deemed at the ACFA rate, for revenue requirement purposes and for the purpose of

calculating credit metrics as part of GCOC proceedings.311

230. Mr. Thygesen submitted that the Commission should use the ACFA funding rates as the

deemed rate for EPCOR’s debt, even if EPCOR does not have access to such funding. He also

stated that any credit metric calculations for EPCOR should be done with the assumption that its

debt is funded through ACFA.312 Mr. Thygesen took note of Decision 2008-100313 in which the

Commission stated, “With respect to a stand alone utility, the directors and management have

responsibilities to ratepayers that include the following: … Accessing the lowest cost financing

at the best terms available to finance utility operations …”314 He submitted that it is clear that the

City of Edmonton and EUI are not accessing the lowest cost financing at the best terms available

for EPCOR, because the lowest cost financing would be funding from ACFA, which is contrary

to the Commission’s direction in Decision 2008-100.315

231. EPCOR contended that the submissions of Mr. Madsen and Mr. Thygesen are outside the

scope of this proceeding. It contended there is no principled basis for the availability of ACFA

financing (or the lack thereof) to have any effect on EPCOR’s ROE, deemed equity ratio, credit

metrics or risk profile.316

232. EPCOR submitted that the evidence of Mr. Madsen and Mr. Thygesen is entirely at odds

with the stand-alone principle. It argued that while Mr. Thygesen purports to rely on previous

Commission findings relating to the stand-alone principle in another context in support of his

309 Exhibit 22570-X0131, PDF page 107. 310 Exhibit 22570-X0888, paragraph 381. 311 Exhibit 22570-X0557, paragraphs 212-213. 312 Exhibit 22570-X0551, paragraph 19. 313 Decision 2008-100: ATCO Electric Ltd. Stand Alone Study, Proceeding 18, Application 1562230-1,

October 21, 2008. 314 Decision 2008-100, page 6. 315 Exhibit 22570-X0551, paragraphs 17-18. 316 Exhibit 22570-X0893, paragraph 76.



position, it is clear there is no reasonable basis to suggest that the stand-alone principle be

abandoned in the present circumstances.317 EPCOR noted Mr. Thygesen’s submission during the

oral hearing that, if the phrase about the owner accessing the lowest cost financing was not in

Decision 2008-100, his view with respect to ENMAX and EPCOR would be that the stand-alone

principle should be considered in order to meet the fair return standard.318

233. EPCOR submitted that the responsibility described in the phrase relied upon by

Mr. Thygesen from Decision 2008-100 is not properly interpreted as an absolute obligation that

overrides the stand-alone principle. It further indicated that the passage does not appear to have

been treated by the Commission as an authoritative statement or principle noting that Decision

2008-100 was not applied by the Commission in the context of addressing the ACFA financing

issue with respect to EPCOR in decisions that were issued subsequent to Decision 2008-100.319

234. EPCOR submitted that the duty of a utility to access the lowest cost financing at the best

terms available to finance utility operations is inconsistent with a previous observation of the

Commission that historically, the least cost approach has not necessarily been accepted when

assessing the prudence of utility costs.320

235. Mr. Bell stated that it is both unfair and unreasonable for a shareholder, who seeks to

maximize its ROE from a regulated utility, not to seek to minimize borrowing costs to the utility

when it had an opportunity to do so.321 The UCA submitted that any reasonable and prudent

shareholder should absolutely look to secure the lowest financing available to its company in the

marketplace.322

236. The UCA submitted that the Commission can and should decline to apply the stand-alone

principle by rote and should decline to apply the stand-alone principle to the detriment of

customers. It indicated that customers are paying an additional $6 million per year as a result of

EPCOR not accessing ACFA financing. The UCA stated that in the circumstances, it is both

reasonable and fair that the Commission deem ACFA funding rates for EPCOR’s debt.323

Commission findings

237. The Commission agrees with the CCA’s recommendation that “If a utility does not avail

itself of ACFA funding (EDTI), and yet that utility has depressed credit metrics as a result, the

Commission should not award that utility additional equity thickness simply on the basis of the

weakened credit metrics resulting from not using ACFA funding.”324 In line with this finding, in

Section 9 the Commission considers the City of Edmonton’s refusal to make ACFA financing

available to EPCOR in its consideration of EPCOR’s credit metrics for the purposes of assessing

the capital structure necessary to provide EPCOR with a fair return.

317 Exhibit 22570-X0733, A42. 318 Transcript, Volume 8, page 1699. 319 Exhibit 22570-X0893, paragraphs 109-110. 320 Exhibit 22570-X0893, paragraph 111. 321 Exhibit 22570-0675, UCA-AUC-2018JAN26-029. 322 Exhibit 22570-X0767.01, paragraph 324. 323 Exhibit 22570-X0913, paragraph 188. 324 Exhibit 22570-X0888, paragraph 381.



238. With respect to EPCOR’s cost of debt and whether it should be deemed at ACFA rates,

the Commission finds that this issue is best determined in a general tariff application (GTA) or

other rate-related proceeding. Accordingly, concerns advanced by interveners in this proceeding

with respect to EPCOR’s cost of debt being higher than necessary given that its parent has access

to ACFA financing will not be addressed in this decision.

7.3 Use of equity funding riders

239. ENMAX provided the following summary of equity funding riders:

Section 138 of the Electric Utilities Act states that a municipality may impose amounts in

respect of its electric distribution system that are in addition to the rates approved by the

Commission, if the bills submitted to customers (a) clearly distinguish between the rates

approved by the Commission and the additional amounts imposed by the municipality,

and (b) identify the additional amounts imposed by the municipality as a surcharge or

tax.325

240. ENMAX noted that historically, the Commission has treated any funds received through

equity funding riders as no-cost capital, because this would mitigate any concerns about double

recovery of investment and an unfair return on investment.326 Mr. Coyne commented that because

of this no-cost capital treatment, any equity funding riders should not enter into consideration

when setting the approved ROE.

241. Similar to his submission on ACFA financing, Mr. Hevert submitted that because

EPCOR does not use equity funding riders, any issues related to its risk profile arising from the

use of equity funding riders are moot. He stated that even if such financing was available,

empirical research indicates that the use of equity funding riders has no statistically significant

effect on required return. Mr. Hevert stated that any use of an equity funding rider would have no

effect on EPCOR’s risk profile.327

242. EPCOR noted that the City of Edmonton has never used an equity funding rider for

EPCOR, and EPCOR has never requested that the city do so.328

Commission findings

243. The Commission finds that the use of an equity funding rider, all else equal, may provide

a municipally owned utility with additional cash flow and could provide a municipally owned

utility with support that would not be available to a non-municipal regulated utility. However,

given that neither Calgary nor the City of Edmonton impose an equity funding rider, the

Commission does not consider it necessary to address this matter any further at this time. The

Commission may revisit the impact and treatment of equity funding riders in a future proceeding

if equity funding riders are implemented.

325 Exhibit 22570-X0129, paragraph 13. 326 Exhibit 22570-X0129, paragraphs 14-15. 327 Exhibit 22570-X0153.01, PDF pages 128-130. 328 Exhibit 22570-X0195, paragraph 54.



8 Return on equity

244. A GCOC proceeding establishes the deemed return on equity for the purposes of setting

regulated rates for a future period; in this case, for 2018, 2019 and 2020. Generally, the cost of

equity to a firm is the return that investors require to make on equity investment in the firm. That

is, investors will only provide funds if the ROE that they expect to receive is sufficient to

compensate them for the risks they are assuming in making the investment. The approved cost of

equity in the GCOC period is a point estimator of investor return expectations that reflects

investors’ return requirements over the long run. In reality, due to the long-term nature of equity

returns, investors evaluate their equity investment both in the short term and the long term, as

actual returns fluctuate over time.

245. The Commission received a significant body of evidence to assist it in determining a fair

approved ROE, including a number of opinions on the proper methodology to be employed and a

wide range of proposed ROEs, based on evidence on the current financial environment and the

results of a number of models.

246. Dr. Cleary, Dr. Villadsen, Mr. Buttke, Mr. Coyne, Mr. Hevert and Mr. Thygesen

provided evidence on changes in the global and Canadian financial environment since the

conclusion of the 2016 GCOC proceeding. The Commission’s findings on this evidence are set

out in Section 6.7.

247. Dr. Cleary, Dr. Villadsen, Mr. Coyne and Mr. Hevert utilized the capital asset pricing

model (CAPM). Dr. Villadsen and Mr. Hevert also employed the use of an empirical CAPM

(ECAPM). The Commission’s findings on the evidence relating to these models are set out in

Section 8.2.4 and Section 8.2.5.

248. Dr. Cleary and Mr. Hevert used a bond yield plus risk premium model (BYPRPM).

Mr. Hevert also utilized a predictive risk premium model. The Commission’s findings with

respect to these models are set out in Section 8.3.

249. Dr. Cleary, Dr. Villadsen, Mr. Coyne and Mr. Hevert submitted discounted cash flow

(DCF) model estimates for utility equities, in order to estimate the required ROE for the affected

utilities. Dr. Cleary and Mr. Hevert submitted DCF model estimates for the Canadian market as a

whole, while Mr. Hevert also submitted a DCF model estimate for the U.S. market as a whole.

The Commission’s findings on the various DCF model results are set out in Section 8.4.

250. Dr. Cleary submitted evidence on the stock market return expectations of finance

professionals. The Commission addresses this area in Section 8.5.

251. Dr. Villadsen presented information on the approved ROE for other Canadian and U.S.

utilities for 2016 and 2017. Dr. Cleary presented evidence on the relevance of market price-to-

book (P/B) values in assessing the cost of equity. The Commission’s findings on this material are

set out in Section 8.7.

252. Consistent with the approach adopted in previous GCOC decisions, in arriving at the fair

approved return for the affected utilities, the Commission considered a variety of approaches,

models and directional indices. The Commission summarizes its findings and sets out the

approved ROE for 2018-2020 in Section 8.8.



253. In Section 8.9, the Commission considered its previous approach of using an annual

adjustment formula for ROE and indicates its intention to explore the possibility of returning to a

formula-based approach to cost of capital matters

8.1 Use of proxy group companies

254. Before the Commission begins its review of the evidence on the financial models

employed by the various witnesses, it will comment on the use of data from proxy group

companies. Dr. Villadsen explained the need for the use of this data, as follows.

Since the Utilities [the ATCO Utilities and AltaGas] are subsidiaries of consolidated

entities and do not themselves have publicly traded stock, it is not possible to directly

estimate their cost of equity using the CAPM or DCF models. This is because these

models rely on market information (such as stock prices, betas based on historical stock

returns, and growth rate estimates) to estimate the expected returns required by equity

investors.329

….

That is why I develop samples of publicly traded companies that are as analogous as

possible to the Utilities in terms of business risk, and apply the models to those samples

as proxies for the Utilities.330

255. Mr. Coyne331 and Mr. Hevert332 echoed the views of Dr. Villadsen.

256. The Commission has previously commented on some of the challenges associated with

determining the ROE for the affected utilities, because of the lack of direct market evidence.

…the determination of the rate of return on equity for a regulated utility is difficult given

that the correct answer is not readily apparent. This determination requires an expert

tribunal to apply its judgment in assessing often conflicting evidence and to consider the

differing interests and perspectives on risk of debt and equity investors. This exercise is

made even more complex in Canada, and in Alberta in particular, given the limited

number of stand-alone utilities issuing debt and the lack of any utilities that issue equity

directly to investors. This fact which has partially resulted from deregulation and

unbundling of utility services, corporate reorganizations creating utility holding

companies, holding companies owning a mix of regulated and unregulated business and

utility acquisitions was referred to in the oral hearing as interposing a “dirty window”

between direct market evidence on cost of capital and the true cost of capital for Alberta

utilities.333

257. Mr. Hevert developed two proxy groups. The first group, referred to as his Canadian

utility proxy group, consists of six publicly traded Canadian utility companies.334 Mr. Hevert’s

second proxy group, referred to as his U.S. utility proxy group, consists of 25 publicly traded

329 Exhibit 22570-X0193.01, A39. 330 Exhibit 22570-X0193.01, A39. 331 Exhibit 22570-X0131, PDF page 33. 332 Exhibit 22570-X0153.01, PDF page 44. 333 Decision 2009-216, paragraph 110. 334 Exhibit 22570-X0153.01, Table 2.



U.S. companies335 that form part of the universe of companies classified by Value Line as electric

utilities.336

258. Mr. Coyne developed three proxy groups. The first group, referred to as his Canadian

utility proxy group, consists of five publicly traded Canadian utility companies.337 The second

proxy group selected by Mr. Coyne, referred to as his U.S. electric proxy group, consists of 11

publicly traded U.S. companies338 that form part of the universe of companies classified by Value

Line as electric utilities.339 Mr. Coyne suggested that these 11 companies would be considered by

investors as comparable in risk to Alberta’s electric utilities.340 Mr. Coyne’s third proxy group,

referred to as his North American electric proxy group, consists of the 11 companies from his

U.S. electric proxy group, and three companies from his Canadian utility proxy group.

Mr. Coyne indicated that the three Canadian companies included in his third proxy group are

primarily engaged in the provision of electricity.341

259. Dr. Villadsen developed five main proxy groups, and she further developed subsample

proxy groups within three of those main proxy groups. Dr. Villadsen’s first proxy group, referred

to as her Canadian utility proxy group, consists of nine Canadian publicly traded companies342

that have utility operations in Canadian regulatory jurisdictions.343 Dr. Villadsen’s second proxy

group, referred to as her U.S. electric utility proxy group, consists of 30 publicly traded U.S.

companies,344 whose primary source of revenues and majority of assets are in the regulated

portion of the U.S. electricity industry.345 Dr. Villadsen developed a subsample from within her

second proxy group, referred to as her subsample U.S. electric utility proxy group. This

subsample consists of 21 companies,346 each of which has at least 80 per cent of their assets

subject to regulation.347

260. Dr. Villadsen’s third proxy group, referred to as her U.S. gas LDC (local distribution

company) utility proxy group, consists of nine publicly traded U.S. companies348 that have the

majority of their revenue generating assets dedicated to the regulated distribution of natural gas

in the U.S.349 Dr. Villadsen developed a subsample from within her third proxy group, referred to

as her subsample U.S. gas LDC utility proxy group. This subsample, which consists of six

companies,350 was developed in order to exclude three companies from her U.S. gas LDC utility

proxy group that are the subject of major mergers and acquisitions.351

335 Exhibit 22570-X0153.01, Table 3. 336 Exhibit 22570-X0153.01, PDF page 45. 337 Exhibit 22570-X0131, Table 4. 338 Exhibit 22570-X0131, Table 5. 339 Exhibit 22570-X0131, PDF page 36. 340 Exhibit 22570-X0131, PDF page 36. 341 Exhibit 22570-X0131, PDF page 38. 342 Exhibit 22570-X0193.01, Figure 8. 343 Exhibit 22570-X0193.01, A41. 344 Exhibit 22570-X0193.01, Figure 9. 345 Exhibit 22570-X0193.01, A46. 346 Exhibit 22570-X0193.01, Figure 9. 347 Exhibit 22570-X0193.01, A43. 348 Exhibit 22570-X0193.01, Figure 10. 349 Exhibit 22570-X1093.01, A47. 350 Exhibit 22570-X0193.01, Figure 10. 351 Exhibit 22570-X1093.01, A47.



261. Dr. Villadsen’s fourth proxy group, referred to as her U.S. water utility proxy group,

consists of eight publicly traded U.S. companies,352 whose primary source of revenues and

majority of assets are subject to regulation.353

262. Dr. Villadsen’s fifth proxy group, referred to as her U.S. pipeline proxy group, consists of

six publicly traded U.S. companies354 that operate primarily in the regulated transportation of

natural gas, crude oil or petroleum products in the U.S.355 Dr. Villadsen developed a subsample

from within her fifth proxy group, referred to as her subsample U.S. pipeline proxy group. This

subsample, which consists of three companies,356 reflects the companies within the U.S. pipeline

proxy group that have a higher proportion of regulated assets dedicated to pipeline transportation

operations.357

263. Dr. Cleary utilized three proxy groups. His first proxy group, referred to as his nine

company Canadian utility proxy group, consists of nine publicly traded Canadian utility

companies.358 Dr. Cleary’s second proxy group, referred to as his seven company Canadian

utility proxy group, is a subsample of his nine company proxy group, and consists of seven

companies.359 Dr. Cleary’s seven company Canadian utility proxy group was developed in order

to exclude two companies that are primarily non-regulated utilities.360

264. Dr. Cleary’s third proxy group, referred to as his four company Canadian utility proxy

group, is also a subsample of his nine company proxy group, and consists of four companies.361

Dr. Cleary’s four company Canadian utility proxy group was developed in order to exclude two

companies that are primarily non-regulated utilities, to exclude two holding companies that

include interests in non-regulated assets, and to exclude one company that has a mix of regulated

and non-regulated assets.362

265. Based on some quantitative analysis, described in more detail in Section 9.3.3, Dr. Cleary

submitted that U.S. holding companies are poor comparators for the affected utilities, because

the U.S. utilities have “significantly higher business risk.”363 Given Dr. Cleary’s view that there

are significant issues with using U.S. companies as proxy groups, Dr. Cleary only used data from

Canadian companies in his CAPM and DCF analysis.364

266. The UCA submitted there is substantial and compelling evidence to support disregarding

and excluding U.S. proxy groups for the purposes of estimating the cost of equity. It further

submitted that if the Commission does not agree that all the U.S. proxy groups should be

352 Exhibit 22570-X0193.01, Figure 11. 353 Exhibit 22570-X1093.01, A48. 354 Exhibit 22570-X0193.01, Figure 12. 355 Exhibit 22570-X1093.01, A49. 356 Exhibit 22570-X0193.01, Figure 12. 357 Exhibit 22570-X0193.01, A49. 358 Exhibit 22570-X0562.01, Table 8. 359 Exhibit 22570-X0562.01, Table 8. 360 Exhibit 22570-X0562.01, PDF page 46. 361 Exhibit 22570-X0562.01, Table 8. 362 Exhibit 22570-X0562.01, PDF page 47. 363 Exhibit 22570-X0562.01, PDF page 47. 364 Exhibit 22570-X0562.01, PDF page 92.



excluded, then Dr. Villadsen’s U.S. pipeline proxy group and U.S. water utility proxy group

should be specifically excluded.365

267. The UCA noted that the beta (β) of Dr. Villadsen’s U.S. pipeline proxy group is 1.04,

while Canadian utility betas do not approach 1.00.366 The UCA suggested that Mr. Hevert and

Mr. Coyne appeared to agree that the U.S. pipeline proxy group should be disregarded.367 The

UCA submitted that if the U.S. pipeline proxy group is not comparable, as acknowledged by

Dr. Carpenter,368 then it should not be used to draw conclusions as to the cost of equity, even as

a directional indicator.369

268. AltaGas and the ATCO Utilities replied that despite Dr. Cleary’s preference for focusing

on regulated entities, the UCA wanted the Commission to exclude Dr. Villadsen’s U.S. water

utility proxy group. They pointed out that the average percentage of regulated assets for the U.S.

water utility proxy group is 94 per cent, and Dr. Carpenter described this sample as “about as

clean a pure play sample as you’re going to find in the regulated utility space in North

America.”370 AltaGas and the ATCO Utilities noted that Dr. Villadsen used her U.S. pipeline

proxy group as an upper bound on the ROE, and her recommended ROE of 10 per cent is lower

than the estimate for her U.S. pipeline proxy group.371

269. As discussed in more detail in Section 9.3.3, Mr. Coyne,372 Dr. Carpenter and

Dr. Villadsen373 all considered that Dr. Cleary’s quantitative-based conclusion that the U.S.

utilities in the various proxy groups have significantly more business risk than the affected

utilities was unsound. They submitted that Dr. Cleary had performed a flawed coefficient of

variation (CV) analysis, and that if Dr. Cleary had performed the correct analysis, he would have

found that U.S. utilities have lower volatility in operating profit margins.

Commission findings

270. The Commission acknowledges the challenges in choosing suitable publicly traded

companies to serve as reasonable comparators to the affected utilities. This is compounded by

the “dirty window” phenomenon as referenced in the above quote from paragraph 110 of the

2009 GCOC decision.

271. As discussed in Section 9.3.3, because of issues identified with Dr. Cleary’s quantitative-

based comparison of the business risks of the affected utilities and U.S. utilities,374 the

Commission is not convinced that there is substantial evidence on which to exclude the use of

U.S. proxy groups.

365 Exhibit 22570-X0897.01, paragraphs 54-55. 366 Exhibit 22570-X0897.01, paragraphs 55-56. 367 Exhibit 22570-X0897.01, paragraphs 55-56. 368 Transcript, Volume 4, page 762. 369 Exhibit 22570-X0888, paragraph 58. 370 Transcript, Volume 1, page 169. 371 Exhibit 22570-X0918, paragraphs 119-122. 372 Exhibit 20622-X0909, PDF page 5. 373 Exhibit 20622-X0909, PDF page 5. 374 Exhibit 22570-X0775, PDF pages 49-51. Exhibit 22570-X0751, A18. Exhibit 22570-X0751, A20.

Exhibit 22570-X0767.01, A19. Exhibit 22570-X0741.01, PDF pages 58-59.



272. Dr. Villadsen’s U.S. pipeline proxy group received a lot of scrutiny as to its

comparability to the affected utilities. Dr. Villadsen stated that she “excluded all my discounted

cash flow numbers from the pipeline sample because they were very high.”375 Dr. Carpenter

acknowledged that the U.S. pipeline proxy group was not comparable to the affected utilities, but

considered it could provide useful information due to the method by which these companies are

regulated.376 Mr. Coyne considered this proxy group to be an outlier to be set aside, when

questioned as to the magnitude of the range of beta estimates provided to the Commission in this

proceeding range.377

273. The Commission agrees with the overall view that Dr. Villadsen’s U.S. pipeline proxy

group, and its subsample group, are not valid comparators for determining the approved ROE for

the affected utilities. The Commission will therefore disregard any results from these proxy

groups as part of its ROE analysis.

274. The Commission has reviewed the selection process followed by Dr. Cleary,

Dr. Villadsen, Mr. Coyne and Mr. Hevert in arriving at each of their proxy groups. With the

exception of Dr. Villadsen’s U.S. pipeline proxy group and its subsample group, the Commission

considers that the selection processes resulted in reasonable proxy groups for application of the

ROE estimation models. Regarding Dr. Villadsen’s U.S. water utility proxy group, the

Commission finds that there is insufficient evidence to exclude this group, beyond Dr. Cleary’s

submission that he “simply did not feel it was a valid comparator sample.”378

275. The Commission retains its view from the 2016 GCOC decision that although returns

awarded by U.S. regulators cannot be used directly in determining a fair return for Alberta

utilities, it is reasonable to consider the U.S. market returns data given the globalization of the

world economy and integration of North American capital markets.379 Accordingly, the

Commission will consider the market-based results from both the Canadian and U.S. proxy

groups in this decision, with the exception of the results from Dr. Villadsen’s U.S. pipeline proxy

group and its subsample group. Even though the Commission agrees that the proxy selection

processes resulted in reasonable proxy groups for application in the ROE estimation models, the

Commission is mindful of the “dirty window” problem, given that none of the affected utilities

raise capital directly in the equity market. Accordingly, a significant amount of judgment by both

witnesses and the Commission must be applied when interpreting this data to establish the ROE

required by investors in the affected utilities.

8.2 The capital asset pricing model

276. The CAPM approach is broadly based on the principle that investors’ compensation for

the use of their capital must recognize two factors: their foregone time value of money, and any

risk attendant in the investment. The time value of money is represented in CAPM by a

component of the required rate of return that corresponds to a risk-free rate, which is intended to

represent the return an investor would expect to receive for investing capital in a risk-free

security over a comparable time period. The second part of CAPM incorporates an adjustment to

the risk-free rate intended to reflect a premium required to address the return required to

375 Transcript, Volume 4, page 664. 376 Transcript, Volume 4, page 762. 377 Transcript, Volume 5, page 953. 378 Exhibit 22570-X0699, Cleary-ATCO/AUI-2018JAN26-014. 379 Decision 20622-D01-2016, PDF page 72.



compensate for the risk of beyond the risk-free rate, referred to as the market equity risk

premium (MERP), and the beta, which is a measure of how sensitive the subject security’s

required return is relative to changes in overall market returns. Beta is usually derived from an

examination of the past statistical relationship between historical returns for a given security and

the returns of the overall capital market during the same time period. In this way, CAPM

calculates the expected return for a security as the rate of return on a risk-free security plus a risk

premium specific to that security or type of security. In other words, the CAPM formally

assumes that all securities are priced such that the required return on the security is equal to the

risk-free rate plus the security’s beta risk measure times the difference between the required

return on the overall market and the risk-free rate.

277. In general terms, CAPM can be represented by the following formula:

Re = Rf +β[E(Rm)-Rf],

where:

Re is the required return on common equity

Rf is the risk-free rate

β, or beta, measures the sensitivity of a required return of an individual security to

changes in the market return

E(Rm)-Rf is the MERP; i.e., the expected market return E(Rm) minus the risk-free rate,

Rf

278. Evidence supporting proposed ROEs based on an application of CAPM, or variations

thereof, was provided by Mr. Hevert, Dr. Villadsen, Mr. Coyne and Dr. Cleary. As well,

Mr. Thygesen and Mr. Johnson commented on some inputs to the CAPM recommendations

presented in this proceeding. Each CAPM component, and the overall resulting CAPM estimates

for ROE, are addressed in sections 8.2.1 to 8.2.5.

8.2.1 Risk-free rate

279. The CAPM analysis requires an estimate of the risk-free rate. As in previous GCOC

proceedings, parties to this proceeding used yields on long-term government bonds as a proxy

for the risk-free rate in their CAPM analyses.

280. Both Mr. Hevert and Dr. Cleary indicated that they used both the current and expected

measures of the long-term government bond rate in developing their risk-free rate

recommendations, consistent with the approach accepted by the Commission in previous GCOC

decisions.

281. For his Canadian utility proxy group, Mr. Hevert used two estimates of the risk-free rate:

the then-current 30-day average yield on 30-year Canada bonds of 2.37 per cent, as well as the

2018 projected 30-year Canada bond yield of 3.08 per cent from the Royal Bank of Canada

(RBC) Economics Research Financial Markets Monthly. During the hearing, Mr. Hevert

provided an updated RBC report showing that the yield on 30-year Government of Canada

(GOC) bonds was expected to increase from 2.45 per cent at the start of 2018 to 3.30 per cent by



the end of 2019.380 For his U.S. utility proxy group, Mr. Hevert used four estimates of the risk-

free rate: the then-current 30-day average yield on 30-year Treasury bonds of 2.77 per cent; the

2018 projected 30-year Treasury yield of 3.33 per cent; the 2019 projected yield of 4.20 per cent;

and the 2020 projected yield of 4.30 per cent obtained from the Blue Chip Financial forecasts.381

Mr. Hevert explained that he preferred the RBC and Blue Chip Financial reports because they

forecast 30-year government bond yields, whereas Consensus Forecasts by Consensus

Economics provides forecasts for 10-year yields thus necessitating an additional adjustment.382

282. Mr. Johnson presented the RBC report from January 2018, showing the same projections

as contained in Mr. Hevert’s undertaking. Mr. Johnson also pointed out that at the time of the

2016 GCOC proceeding, “RBC was forecasting essentially the exact same 3.30% LTC [long-

term Canada] yield two year’s [sic] out as they are now.”383 As such, Mr. Johnson concluded that

there has been no change in the forecast interest rate environment since 2014.

283. Dr. Cleary presented a risk-free rate range of 2.2 per cent to 3.0 per cent, with a mid-

point of 2.6 per cent. The lower bound of 2.2 per cent represented the rounded-up actual

prevailing long-term Canada yield as of December 2017 when Dr. Cleary prepared his evidence.

The upper bound of 3.0 per cent was obtained by adding the long-term average spread between

10- and 30-year GOC bond yields of 50 basis points to the October 2017 Consensus Forecasts

for GOC 10-year yields of 2.5 per cent for October 2018.384 Dr. Villadsen and Mr. Coyne

criticized Dr. Cleary’s forecast horizon of October 2018 as being too short, given that this

proceeding determines the cost of capital for 2018 to 2020.385

284. Dr. Villadsen expressed the view that because all indicators point to an increase in the

cost of debt going forward, a forecast bond rate is more indicative of the cost of equity than the

current rate. To develop her risk-free estimate, Dr. Villadsen relied on a forecast of the GOC

bond yields in 2019, which is the middle year of the test period for this proceeding. Dr. Villadsen

identified that the October 2017 Consensus Forecasts predicted the 10-year GOC bond yield to

be 2.9 per cent by 2019. To that predicted yield she added 40 bps based on her estimate of the

representative maturity premium for the 30-year over the 10-year GOC bonds, to arrive at a

lower bound of her risk-free rate recommendation of 3.3 per cent. Dr. Villadsen also considered

a scenario in which the risk-free interest rate was 3.45 per cent.386

285. Mr. Coyne expressed a similar preference for using forward-looking data rather than

current risk-free rates. Relying on the October 2017 Consensus Forecasts data for predicted

10-year government bond yields for each of 2018, 2019 and 2020, Mr. Coyne calculated an

average rate for the period of 2.83 per cent for Canada and 3.27 per cent for the U.S. After

adding an average historical spread between 10- and 30-year government bond yields (43 bps for

Canada and 59 bps for the U.S.), Mr. Coyne arrived at the long-term bond yields of 3.26 per cent

for Canada and 3.95 per cent for the U.S.387

380 Exhibit 22570-X0869. 381 Exhibit 22570-X0153.01, PDF pages 75-76. 382 Transcript, Volume 6, page 1183. 383 Exhibit 22570-X0611.02, PDF page 10. 384 Exhibit 20622-X0306, PDF page 31. 385 Exhibit 22570-X0767.01, PDF pages 29-30. Exhibit 22570-X0775, PDF page 24. 386 Exhibit 22570-X0193.01, PDF page 59. 387 Exhibit 22570-X0131, PDF page 47.



286. In his evidence for the CCA, Mr. Thygesen compared the interest rate predictions by

Consensus Forecasts to the actual interest rates and stated that Consensus Forecasts (and other

forecasts by banks and government bodies) “have consistently over-forecast the 10-year rate

since 2008” and therefore exhibit “a strong bias towards over-forecasting.”388 As a result,

Mr. Thygesen argued against relying solely on Consensus Forecasts. In their respective

arguments, the CCA and the UCA supported this view.389

287. Mr. Thygesen reiterated his recommendation from the 2016 GCOC proceeding that the

Commission consider forward curve rates in developing its risk-free estimates; for example, by

taking an average of the Consensus Forecasts and the forward curve rates. Mr. Thygesen

acknowledged that the forward curve rates are not a forecast per se; however, they are based on

market transactions and have had a smaller forecasting error as compared to Consensus Forecasts

over the 2016-2017 period. Based on the utility witnesses’ responses to the CCA IRs,

Mr. Thygesen presented data indicating that forward curve rates for long-term GOC bond yields

were in the 2.3 to 2.6 per cent range, with the majority of data points centered on the 2.3 per cent

estimate.390

288. The utility witnesses disagreed with recommendations to assign less weight to interest

rate forecasts. They indicated that Mr. Thygesen did not perform the statistical analysis required

to demonstrate the presence of bias in economic forecasts. They also pointed out that the post-

financial crisis period referenced by Mr. Thygesen, over which the forecasts were made,

exhibited many unusual characteristics and, as a result, interest rate forecast accuracy was low

during that period.391

289. Regarding the use of forward curve rates, Mr. Buttke stated that “forward curves have not

been proven to be more accurate than forecasts in an academically robust way.” He explained

that the “unbiased expectations theory” underlies the premise that forward rates would provide

an accurate forecast. In this regard, Mr. Buttke referenced a study by the Federal Reserve, which

concluded that because “The expectations hypothesis of the term structure has been consistently

and decisively rejected, for the United States at least, and so we should not expect to find that

forward interest rates and interest rate futures are efficient forecasts of future interest rates.”392

Mr. Buttke also indicated that forward curve rates can be inaccurate because they reflect market

equilibrium (including any market inefficiencies) for a set of facts that is known at a given

moment. In other words, they project whatever is currently known into the future. Because of

this, forward curve rates may be “especially poor at implying future prices when markets are

changing level or direction – they almost always imply a continuation of current conditions and

trends.”393 Mr. Buttke provided charts showing that forward curves “over predict” the status quo:

in an interest rate market that has trended lower for a number of years, the forward curve

388 Exhibit 22570-X0551, PDF pages 9 and 11. 389 Exhibit 22570-X0897.01, paragraph 19. Exhibit 22570-X0888, paragraph 104. 390 Exhibit 22570-X0551, PDF pages 13-14. 391 Exhibit 22570-X0749, PDF page 30. Exhibit 22570-X0193.01, PDF page 31. Exhibit 22570-X0775,

PDF page 15. Exhibit 22570-X0741.01, PDF page 39. 392 Exhibit 22570-X0749, PDF page 31. 393 Exhibit 22570-X0749, PDF page 33.



projections will tend to be too high. If a market has a trend higher in rates, forward curves will

tend to be too low.394

290. At the same time, Mr. Buttke reiterated his statements from the 2016 GCOC proceeding

that “forward rates are a data point – they do not necessarily have to be ignored completely, but

their influence as an input should be weighted accordingly.”395 Other utility witnesses came to a

similar conclusion that caution needs to be exercised when relying on forward rates in

developing the forecasts. Mr. Coyne cautioned “interpreting this data is complicated and would

not be a transparent input to the regulatory process.”396 Mr. Hevert pointed out that forward yields

have been quite volatile in the period leading up to this proceeding; however, despite the

volatility, “they consistently have indicated expectations for interest rate increases.”397 Both

Mr. Hevert and Dr. Villadsen indicated that, in any event, implied forward curve rates are well

known and considered by professionals making forecasts, such as Consensus Forecasts.398

291. As mentioned in Section 6, Mr. Thygesen also drew the Commission’s attention to the

flattening of the yield curve. Mr. Thygesen referenced several articles indicating that the U.S.

yield curve is flattening with the difference between short-term and long-term yields being at its

lowest since November 2007.399 The articles also indicated that if the yield curve becomes

inverted (with long-term rates below short-term rates), this “has proven a reliable indicator of

impending economic slumps.”400

292. Mr. Buttke countered that it is not always the case that a flattening yield curve may

become inverted thus signalling the advent of lower interest rates, and it is important to know

what drives the shape of the yield curve. In this regard, Mr. Buttke pointed to the same

Bloomberg article cited by Mr. Thygesen, as well as U.S. Treasury press releases, which led

Mr. Buttke to conclude that changes in the mix of Treasury bonds to include a greater proportion

of notes in two- to five-year maturities has changed and influenced the shape of the yield

curve.401 Mr. Buttke pointed out that inverted yield curves are typically associated with restrictive

monetary policy,402 and he presented charts showing that “yield curves have gone through many

periods where they have flattened only to re-steepen quickly or to remain flat for a number of

years and then re-steepen.”403 Based on the above, Mr. Buttke concluded that “There is no reason

to assume that current yield curve levels are outside of historical ranges and little reason to

predict that a recession is likely to happen in the near term based on the shape of the yield

curve.”404

394 Exhibit 22570-X0749, PDF page 35. 395 Exhibit 22570-X0749, PDF page 39. 396 Exhibit 22570-X0775, PDF page 18. 397 Exhibit 22570-X0741.01, PDF page 37. 398 Exhibit 22570-X0193.01, PDF page 34. Exhibit 22570-X0741.01, PDF page 41. 399 Exhibit 22570-X0551, PDF pages 20-25. 400 Exhibit 22570-X0551, PDF pages 21-24. 401 Exhibit 22570-749, PDF page 51. 402 Exhibit 22570-749, PDF pages 53-54. 403 Exhibit 22570-749, PDF pages 56-60. 404 Exhibit 22570-749, PDF page 60.



Commission findings

293. In Decision 3539-D01-2015, the Commission considered that “the forward curve acts as

an indication of what future interest rates are currently expected to be and can be considered for

forecasting purposes.”405 The Commission continues to hold this view, while acknowledging the

limitations of relying on the implied forward curve rates, as they are not “necessarily pure

measures of market expectations.”406 In general, the Commission agrees with the view that

implied forward yields are among the data points that can be used to develop interest rate

forecasts.

294. The Commission also continues to see merit in using both the current and expected

interest rates in considering a reasonable risk-free rate forecast. This was the Commission’s

approach in the 2013 and 2016 GCOC decisions and that of Mr. Hevert and Dr. Cleary in this

proceeding. As Dr. Cleary explained, utilizing the existing rates as a forecast is an accepted

method that “offer[s] the benefit of a starting point that reflects actual yields (i.e., yields that

investors can actually achieve today), rather than forecasts which may or may not materialize.”407

This approach has been of assistance to the Commission following the 2008-2009 financial

crisis, when interest rates and other financial indicators behaved in a less-than-predictable way.408

295. As illustrated in Figure 7, over the course of this proceeding (November 2017 to March

2018), the yield on long-term GOC bonds fluctuated around the 2.3 per cent level, which was

also the average yield for that period.409 The Commission finds this to be a reasonable starting

point for the risk-free rate in its current analysis.

296. Regarding the expected rates, the Commission has examined the long-term rate forecasts

put forward by parties and observes that there appears to be a broad consensus among various

forecasting bodies (shared by most parties in this proceeding) that long-term interest rates are

likely to rise throughout the 2018-2020 period. Nevertheless, the pace and magnitude of any

increase remain uncertain given the evidence. For example, the RBC Financial Markets Monthly,

relied upon by Mr. Hevert and Mr. Johnson, predict long-term government bond yields to reach

3.3 per cent in Canada and 3.85 per cent in the U.S. by the end of 2019. However, the

Commission notes that the RBC reports from September 2017 and October 2017 predicted the

GOC long-term rates to reach 3.3 per cent by the end of 2018,410 but starting in January 2018,

RBC revised these forecasts downwards with rates predicted to be 3.15 per cent by the end of

2018 and reaching 3.30 per cent in the second half of 2019.411 The U.S. forecasts were similarly

revised.

297. Also potentially tempering forecasted interest rate increases is the flattening yield curve.

In Section 6, the Commission took note of the flattening yield curve for Canadian and U.S.

government bond yields experienced in the period leading up to, and over the course of this

405 Decision 3539-D01-2015, paragraph 834. 406 Exhibit 22570-X0749, PDF page 31. 407 Exhibit 22570-X0562.01, PDF page 10. 408 For example, the Commission observed in the 2016 GCOC decision that, rather than increasing to just under

four per cent, as was generally expected in the 2013 GCOC proceeding, yields on long-term GOC bonds fell by

some 100 bps, from approximately 3.0 per cent to approximately 2.0 per cent. 409 Exhibits 22570-X0835 and 22570-X0836. 410 Exhibit 22570-X0159, PDF pages 92 and 101. 411 Exhibit 22570-X0869.



proceeding, as depicted in Figure 1. Taking into account the current and expected flattening of

the yield curve, the Commission concludes that over the 2018-2020 test period, the spread

between 10-year and 30-year GOC bonds is likely to be lower than the historical average of

some 50 bps that the Commission has accepted in past GCOC decisions. All other things being

equal, this calls for lower long-term estimates derived from the Consensus Forecasts, which only

predicts yields on 10-year GOC bonds. As well, the Commission finds that this flattening of the

yield curve may imply that long-term interest rates may not rise in lockstep, or at all, with the

increase in the short-term rates.

298. Mr. Buttke expressed his view that the driving forces behind the current flattening of the

yield curve give “little reason to predict that a recession is likely to happen in the near term based

on the shape of the yield curve”412 and presented charts showing that “yield curves have gone

through many periods where they have flattened only to re-steepen quickly or to remain flat for a

number of years and then re-steepen,”413 and as such, “a flattening curve does not mean that bond

yields cannot rise.”414 In contrast, Mr. Thygesen referenced publications claiming that a flattening

yield curve argues against higher interest rates and that if the yield curve does become inverted,

this historically has been a reliable precursor of recessions with the yield curve inverting just

before each of the past seven American recessions.415 In the Commission’s view, it is not possible

to discount the likelihood of either outcome at this time (i.e., that the flattening yield curve may

steepen or remain flat with long-term rates still increasing, as surmised by Mr. Buttke, or that the

flattening of the yield curve will continue until it inverts which, in the past, has been an indicator

of a pending recession).

299. In light of the above and considering the findings in Section 6, the Commission cannot

reasonably conclude that the long-term interest rates (as measured by the yield on long-term

Canada bonds) are likely to increase significantly, if at all, over the 2018-2020 test period.

Accordingly, the Commission finds that the prevailing yield on long-term GOC bonds of 2.3 per

cent represents a reasonable estimate of the risk-free rate over the 2018-2020 term. In the

Commission’s view, it is reasonable to expect some continued fluctuation in long-term interest

rates, both upward and downward, around the 2.3 per cent estimate over the forecast period.

8.2.2 Market equity risk premium

300. Dr. Villadsen provided the following description of the MERP:

Like the cost of capital itself, the market equity risk premium is a forward-looking

concept. It is by definition the premium above the risk-free interest rate that investors can

expect to earn by investing in a value-weighted portfolio of all risky investments in the

market. The premium is not directly observable, and must be inferred or forecasted based

on known market information.

One commonly use [sic] method for estimating the MERP is to measure the historical

average premium of market returns over the income returns on risk-free government

bonds over some long historical period.

412 Exhibit 22570-X0749, A35. 413 Exhibit 22570-X0749, A35. 414 Exhibit 22570-X0749, A35. 415 Exhibit 22570-X0551, PDF page 21.



An alternative approach to estimating the MERP eschews historical averages in favor of

using current market information and forecasts to infer the expected return on the market

as a whole, which can then be compared to prevailing government bond yields to

estimate the equity risk premium.416

301. Dr. Villadsen used the arithmetic average of annual observed Canadian MERPs from

1935 to the present as her historical MERP, and used the resulting figure of 5.7 per cent as the

MERP in all of her CAPM scenario one calculations.417 Mr. Coyne used an arithmetic average for

his historical Canadian MERP. Using data from 1919 to 2016, he reported a result of 5.60 per

cent. Using data from 1926 to 2016 for the U.S., Mr. Coyne reported an arithmetic average for

U.S. MERP of 6.94 per cent.418 Dr. Cleary reported historical Canadian MERPs from 1900

to 2015. Using the arithmetic average, the result was 5.2 per cent. Using the geometric average,

the result was 3.3 per cent.419

302. Dr. Villadsen, Mr. Coyne and Mr. Hevert also derived forward-looking expected MERP

values.

303. Dr. Villadsen provided expected market return rates for Canada and the U.S., determined

by Bloomberg using a multi-stage dividend discount model. She also provided the expected

10-year risk-free rates that Bloomberg deduced from the expected market return rates to arrive at

their forward-looking MERPs. Dr. Villadsen made a further reduction in order to reflect the

spread between the 10-year risk-free rates and the 30-year risk-free rates. The results were a

forward-looking MERP estimate of 9.49 per cent for Canada, and 6.76 per cent for the U.S.420

304. Based on her proposal that investors’ level of risk aversion remains elevated, and the

forward-looking MERP estimates being higher than the average, Dr. Villadsen used 8.00 per cent

as the MERP in her CAPM scenario two calculations. She stated that this figure is between the

forward-looking MERP estimates of 6.76 per cent for the U.S. and 9.49 per cent for Canada.

Dr. Villadsen justified the use of Canadian and U.S. MERP estimates because of the substantial

interaction of the two markets.421

305. Mr. Coyne argued that since both the U.S. and Canadian economies have enjoyed a

prolonged low interest rate environment, it should be expected that the historical arithmetic

average will understate the current market risk premium.422 Consequently, he incorporated a

forward-looking MERP estimate to respond to changes in capital market conditions.

306. Applying a single-stage DCF methodology, Mr. Coyne calculated the expected market

return rates for Canada and the U.S. on a market capitalization-weighted basis for the individual

companies in each broad market index (the S&P 500 index for the U.S. and the S&P/TSX

Composite index for Canada). He then subtracted his recommended risk-free rates from the

416 Exhibit 22570-X0192.01, PDF pages 24-25. 417 Exhibit 22570-X0193.01, A56. 418 Exhibit 22570-X0131, PDF page 55. 419 Exhibit 22570-X0562.01, Figure 10. 420 Exhibit 22570-X0193.01, A29. 421 Exhibit 22570-X0193.01, A56. 422 Exhibit 22570-X0131, PDF page 56.



expected market return rates to arrive at a MERP estimate of 9.38 per cent for Canada, and

8.89 per cent for the U.S.423 The results are shown in the following table.

Table 2. Forward-looking expected MERPs as reported by Mr. Coyne424

Canada U.S.

%

Expected market return rates 12.64 12.74

Deduct: recommended 30-year risk-free rates 3.26 3.85

Forward-looking expected MERPs 9.38 8.89

307. Noting that the Canadian and U.S. markets are highly correlated, Mr. Coyne averaged the

historical and forward-looking MERP estimates for Canada and the U.S., to arrive at a MERP

value of 7.70 per cent,425 which he used in his CAPM analysis.426

308. Mr. Hevert proposed that it is important to ensure the expected market return rates and

the associated MERP are prospective in nature.427 He derived forward-looking expected MERPs

for Canada and the U.S. using two methods. The first method calculated the expected return rates

on the Canadian and U.S. markets, based on the constant growth DCF model, using data

provided by Bloomberg. He then subtracted the actual 30-year risk-free rates and calculated the

results. He also subtracted the projected 30-year risk-free rates and calculated the results.

309. Mr. Hevert’s second method included a semi-log form regression-derived estimate, which

used monthly historical returns on the Canadian and U.S. stock markets as the dependent

variables, relative to monthly historical yields on long-term government bonds as the

independent variables. He then applied the obtained regression coefficients to the actual and

projected 30-year risk-free rates.428

310. The results of Mr. Hevert’s two methods are set out in the following table.

423 Exhibit 22570-X0131, PDF page 56. 424 Exhibit 22570-X0132, worksheet JMC-3 Canada MRP. Exhibit 22570-X0132, worksheet JMC-4 U.S. MRP. 425 Historical Canadian of 5.60 per cent. Historical U.S. of 6.94 per cent. Forward-looking Canadian of 9.38 per

cent. Forward-looking U.S. of 8.89 per cent. Average of these four is 7.70 per cent. 426 Exhibit 22570-X0131, PDF pages 57-58. 427 Exhibit 22570-X0153.01, PDF page 86. 428 Exhibit 22570-X0153.01, PDF pages 86-87.



Table 3. Forward-looking expected MERPs as reported by Mr. Hevert429

Canada Canada U.S. U.S.

(%)

Method 1

Expected market return rates 14.84 14.84 13.83 13.83

Deduct: actual 30-year risk-free rates 2.37 2.77

Deduct: projected 30-year risk-free rates 3.01 3.30

Forward-looking expected MERPs 12.47 11.83 11.06 10.53

Method 2

Regression applied to actual 30-year risk-free rates 6.89 9.74

Regression applied to projected 30-year risk-free rates 5.37 8.77

311. By averaging the results of the two methods for the Canadian market, Mr. Hevert derived

his recommended forward-looking expected MERP of 9.14 per cent for his Canadian sample. By

averaging the results of the two methods for the U.S. market, Mr. Hevert derived his

recommended forward-looking expected MERP of 10.02 per cent for his U.S. sample.430

312. Dr. Cleary indicated that it is common practice to use a range of 3-7 per cent for the

MERP when using the CAPM, with the large majority of MERP estimates falling in the 4-6 per

cent range.431 He provided evidence which he claimed verified that a well-respected finance

professional, textbook author, and provider of financial data uses MERPs in the 4-6 per cent

range, and varies the choice of MERP to reflect the level of uncertainty in the market.432

313. Based on his belief that stock markets reflect fairly normal conditions, but are

experiencing below average volatility, Dr. Cleary used a MERP of five per cent. He stated that

this is the mid-point of the commonly used 4-6 per cent range, and it is 20 bps below the long-

term historical arithmetic average Canadian MERP of 5.2 per cent.433 Dr. Cleary added his

recommended MERP of five per cent to his recommended risk-free rate of 2.6 per cent, and

noted that the resulting 7.6 per cent figure is consistent with his point estimate of 7.5 per cent for

the expected long-term Canadian stock market return rate.434

314. Dr. Cleary stated that the forward-looking expected MERPs reported by Mr. Hevert and

Mr. Coyne were derived based on analyst estimates of growth rates that far exceed GDP growth.

He suggested that Dr. Villadsen’s forward-looking expected MERP suffered from the same

shortcoming.435

315. Mr. Hevert and Mr. Coyne submitted that Dr. Cleary’s partial reliance on historical

MERP values during the current period of low interest rates will understate the cost of equity

429 Exhibit 22570-X0154.01, worksheets Sch 6 MRP TSX, Sch 6 MRP TSX RA, Sch 6 MRP S&P 500, Sch 6

MRP SBBI RA. 430 Exhibit 22570-X0153.01, PDF page 87. 431 Exhibit 22570-X0562.01, PDF page 38. 432 Exhibit 22570-X0562.01, PDF page 41. 433 Exhibit 22570-X0562.01, PDF page 38. 434 Exhibit 22570-X0562.01, PDF page 35. 435 Exhibit 22570-X0562.01, PDF pages 42-43.



because of the inverse relationship between interest rates and the observed MERPs.436 They

explained that if current interest rates are low relative to historical levels, then the current

MERPs should be relatively higher than their historic levels.437

316. The UCA submitted that other than the regression analysis provided by Mr. Coyne to

demonstrate the relationship between interest rates and observed MERPs, no evidence was

provided on this assumed relationship.438 The UCA noted Mr. Coyne’s concession during the oral

hearing that his regression was not statistically significant, and he did not rely on it.439 Dr. Cleary

did not agree that, in general, the MERP increases as interest rates decrease.440 The UCA argued

that if the relationship does exist, the suggestion of rising interest rates brought forward by the

Alberta utilities would lead to a decrease in the MERP.441

Commission findings

317. The historical Canadian MERP values reported by Dr. Villadsen (5.7 per cent),

Mr. Coyne (5.6 per cent) and Dr. Cleary (5.2 per cent) were all developed using arithmetic

averages. Despite the different time periods used, the MERP values are within a relatively

narrow range. The same cannot be said for the forward-looking expected market return rates that

Dr. Villadsen, Mr. Coyne and Mr. Hevert used for Canada, when compared to Dr. Cleary’s

expected long-term market return rate for Canada.

318. Dr. Villadsen’s expected market return rates for Canada range from 12.79 to 12.94 per

cent using a forward-looking MERP value for Canada of 9.49 per cent, and her risk-free rate

estimates for Canada of 3.30-3.45 per cent. Mr. Coyne’s expected market return rate for Canada

is 12.64 per cent. Mr. Hevert’s expected market return rate for Canada is 14.84 per cent. The

resulting range of the utilities experts is 12.64-14.84 per cent, and the average of the four figures

is 13.30 per cent.442 This contrasts significantly with the 7.5 per cent that Dr. Cleary considers to

be a reasonable point estimate for the expected market return rate for Canada, as described

in Section 8.5.

319. As noted by Dr. Cleary, the expected market return rates used by Mr. Coyne and

Mr. Hevert use analyst estimates of growth rates that far exceed expected GDP growth. The

Commission has commented in Section 8.4 that market return growth rates that far exceed

expected GDP growth are not sustainable, particularly for utilities. The Commission finds that

because Mr. Coyne’s and Mr. Hevert’s proposed market return rates significantly exceed

expected GDP growth rate, these estimates are too high. However, no evidence was provided on

the record that would enable the Commission to quantify the extent of these overstatements.

320. With respect to Dr. Cleary’s point estimate of 7.5 per cent for the expected market return

rate for Canada, the Commission has addressed this in Section 8.5.

436 Exhibit 22570-X0741.01, PDF page 42. Exhibit 22570-X0775, PDF page 25. 437 Transcript, Volume 6, page 1198. Exhibit 22570-X0775, PDF page 25. 438 Exhibit 22570-X0913, paragraph 48. 439 Transcript, Volume 5, page 881. 440 Transcript, Volume 9, page 2005. 441 Exhibit 22570-X0913, paragraph 49. 442 Average of 12.79 per cent, 12.94 per cent, 12.64 per cent and 14.84 per cent.



321. The Commission has been presented with a range of 7.5 to 14.84 per cent for the

expected market return for Canada. The Commission finds this range too wide to be informative.

Directionally, the Commission cannot take any guidance from the changes in the MERP

estimates that were provided in the 2016 GCOC proceeding, because some estimates increased,

some decreased and others remained unchanged.

322. Consequently, the Commission will place no weight on the expected market return rates

for Canada in assessing a reasonable MERP value. As a result, the Commission will consider the

historical Canadian MERP rates on the record of the proceeding, and the results produced by Mr.

Hevert’s regression method, in determining a reasonable MERP.

323. As mentioned above, Dr. Villadsen, Mr. Coyne and Dr. Cleary provided historical MERP

rates for Canada that range from 5.2 per cent to 5.7 per cent. The results of Mr. Hevert’s

regression model are 5.37 per cent using a risk-free rate of 3.01 per cent, and 6.89 per cent using

a risk-free rate of 2.37 per cent. In Section 8.2.1 the Commission found that the prevailing yield

on long-term GOC bonds over the course of this proceeding of 2.3 per cent represents a

reasonable estimate of the risk-free rate over the 2018-2020 term. Using Mr. Hevert’s regression

analysis as a guide, this suggests a MERP that is in excess of 6.89 per cent. The use of a MERP

in excess of 6.89 per cent corresponds to the submissions of Mr. Hevert and Mr. Coyne that, in

the current low interest rate environment, the forward-looking MERP should be greater than the

historical Canadian average, which has ranged from 5.2 to 5.7 per cent. In the 2016 GCOC

decision, the Commission acknowledged the inverse relationship between the risk premium and

the level of interest rates.443 The Commission continues to acknowledge this relationship.

8.2.3 Beta

324. The final element of the CAPM is the beta (β) coefficient. Beta is a statistical measure

describing the relationship of a given security’s return with that of the equity market as a whole.

In essence, beta is a measure of market risk of an equity security. Past data (with or without

adjustment) is normally used to estimate the expected beta going forward. As expressed in

previous GCOC decisions, the Commission considers that the appropriate beta to use is one that

reasonably represents the relative risk of stand-alone Canadian utilities.

325. The betas that Mr. Coyne estimated were 0.75 for his Canadian utility proxy group, 0.67

for his U.S. electric proxy group, and 0.68 for his North American electric proxy group,444 using

the estimates from Value Line and Bloomberg, based on weekly stock returns over a five-year

period. Both beta estimation techniques are adjusted to compensate for the tendency to revert

toward the market mean of 1.0 over time.

326. Dr. Villadsen used adjusted historical betas obtained from Bloomberg, using weekly

returns over a three-year period. In applying her beta calculation, Dr. Villadsen developed value-

weighted portfolio betas for each of her proxy groups, which she explained is warranted since it

may cancel out any idiosyncratic fluctuations of an individual company and provide a better

estimate of beta.445 Dr. Villadsen’s proxy groups yielded the following average betas: Canadian

utility proxy group: 0.850; U.S. electric utility proxy group: 0.614; U.S. gas LDC utility proxy

443 Decision 20622-D01-2016, paragraph 228. 444 Exhibit 22570-X0131, PDF pages 48-53. 445 Exhibit 22570-X0193.01, PDF pages 61-62.



group: 0.669; and U.S. water utility proxy group: 0.750. The value-weighted portfolio betas for

each of her proxy groups were Canadian utility proxy group: 0.950; U.S. electric utility proxy

group: 0.578; U.S. gas LDC utility proxy group: 0.659; and U.S. water utility proxy group:

0.644.

327. Mr. Hevert relied on adjusted beta estimates from Value Line and Bloomberg based on

five years of weekly return data. Mr. Hevert’s resulting average of the adjusted beta estimates

was 0.72 for his Canadian utility proxy group, and 0.62 for his U.S. utility proxy group.446

328. To develop his beta range, Dr. Cleary analyzed betas using total monthly returns for the

TSX Utilities Index for several different periods from 1998 to 2017 and compared this to the

average beta of his three proxy groups as of November 2017, based on 60 months of returns.

Dr. Cleary determined that combining the analysis resulted in a reasonable range of 0.30 to 0.60.

To be consistent with previous proceedings, Dr. Cleary put forth the mid-point of 0.45 as his best

point estimate. Dr. Cleary explained that this is slightly above the long-term average Canadian

utility beta estimate of 0.35.447

329. The utilities pointed out that Dr. Cleary’s reported beta coefficients have significantly

increased, from an average of 0.21 in the 2016 GCOC proceeding to an average of 0.43 in the

current proceeding. The utilities pointed out that Dr. Cleary’s directional increase in beta is

consistent with Mr. Hevert’s findings in this proceeding, and explained that this clearly indicates

that the relative risk of Canadian utilities has increased.448

330. A point of disagreement in this proceeding was whether adjusted or unadjusted betas,

often referred to as “raw betas,” should be used in the CAPM. Adjusted betas refer to betas

derived from adjustments to the raw betas for the purpose of forward estimation. For example,

the “Blume” adjustment (named after Professor Marshall Blume) is a well-known method by

which adjusted betas are calculated by giving two-thirds weight to the calculated raw beta and

one-third weight to the market average beta of one.449

331. Dr. Cleary explained that when developing beta estimates for Canadian utilities, it is

inappropriate to use betas that are adjusted toward 1.0, since they have averaged 0.31-0.35 over

the last 25-28 years, and have never approached 1.0 in practice.450 Dr. Cleary noted Mr. Hevert’s

comment that the purpose of the Blume adjustment is to adjust the beta toward its mean value.

As a result, Dr. Cleary submitted that the utilities’ beta should be adjusted toward its mean value

rather than the market value of 1.0.451 Dr. Cleary explained that the Blume adjustment is not

founded on any conceptual basis, but rather it is purely empirical in nature.

332. Mr. Hevert explained that adjusted betas are commonly used in standard practice and

serve as a means to address the Commission’s concerns with respect to the wide range of betas

provided on the record of the last GCOC proceeding.452 Mr. Hevert compared adjusted and

unadjusted betas for his Canadian utility proxy group and found that the adjusted betas’ variation

446 Exhibit 22570-X0153.01, PDF page 106. 447 Exhibit 22570-X0562.01, PDF pages 44-48. 448 Exhibit 22570-X0890.01, PDF pages 21-22. 449 Exhibit 22570-X0913, PDF page 21 450 Exhibit 22570-X0565, PDF pages 3-4. 451 Exhibit 22570-X0897.01, PDF pages 20-21. 452 Decision 20622-D01-2016, paragraph 317.



for the time period analyzed was much lower. Mr. Hevert explained this suggests that use of

adjusted betas addresses the Commission’s concerns with respect to the wide range of betas.453

In response to Dr. Cleary’s comment that the adjustment should be toward the utilities’ average

beta, Mr. Hevert noted that:

because Blume’s research was based on Beta coefficients estimated relative to the market

as a whole, his correction, which is approximated by an α of 0.67, cannot be translated to

an adjustment to the raw Beta coefficient assuming a non-market mean Beta coefficient,

such as Dr. Cleary’s 0.35 average454

333. Mr. Coyne also considered that Dr. Cleary’s recommended beta of 0.45 should be

dismissed as an outlier, in part because the Blume adjustment was not applied.455

334. Another point of disagreement in this proceeding was whether monthly or weekly betas

should be used to develop CAPM estimates.

335. Dr. Villadsen explained that Dr. Cleary presented only monthly betas and relied nearly

exclusively on those to inform his recommendation. Dr. Villadsen submitted that Dr. Cleary

ignores the fact that using weekly data is statistically superior relative to monthly betas during

the time period which he analyzes.456 Dr. Villadsen explained that monthly betas “have become

statistically imprecise and unreliable in the years following the global financial crisis”457 and

weekly betas have become the “standard practice.”458 Dr. Villadsen compared unadjusted weekly

and monthly betas and found that the estimation of error was approximately twice as large using

monthly data.459 Dr. Villadsen explained further that Dr. Cleary’s long-term beta estimate is

biased downward since it considers anomalous periods such as the dot-com bubble period.460

336. Mr. Hevert explained that weekly data as opposed to monthly data is more appropriate

because monthly data gives less weight to the market movements experienced over shorter time

periods and, as a result dampens volatility.461 Additionally, Mr. Hevert compared monthly and

weekly betas of his Canadian utility proxy group and his U.S. utility proxy group, and found that

there are a greater number of negative beta coefficients observed when monthly returns are

assumed.462

337. Mr. Coyne presented several charts in order to compare monthly and weekly betas and

commented that the use of weekly returns tends to correlate more closely with the market than do

monthly returns. Mr. Coyne conducted a statistical analysis to determine the explanatory power

of weekly and monthly beta coefficients. Observing the results of his analysis, Mr. Coyne noted

that weekly betas were statistically significant over the two- and five-year periods analyzed,

453 Exhibit 22570-X0153.01, PDF page 80. 454 Exhibit 22570-X0890.01, PDF page 52. 455 Transcript, Volume 5, page 954. 456 Exhibit 22570-X0767.01, PDF pages 39-40. 457 Exhibit 22570-X0767.01, PDF page 47. 458 Transcript, Volume 2, page 288. 459 Exhibit 22570-X0767.01, PDF pages 128-129. 460 Exhibit 22570-X0767.01, PDF page 135. 461 Exhibit 22570-X0153.01, PDF page 81. 462 Exhibit 22570-X0153.01, PDF page 83.



whereas the monthly betas were not.463 In his evidence, Mr. Coyne recognized that both monthly

and weekly returns are commonly accepted in practice; however, due to the results of his

analysis, Mr. Coyne recommended weekly five-year or two-year betas.464

338. Dr. Cleary submitted that both monthly and weekly return data are widely used to

determine beta coefficients. Dr. Cleary further explained that he could not offer a definitive

opinion on whether monthly or weekly data is best to calculate betas and he did not think that the

Commission could conclusively decide the issue.465 The UCA submitted that the pragmatic

approach is to consider both monthly and weekly return data, which is the approach adopted by

Dr. Cleary.466

339. The UCA disagreed with Dr. Villadsen’s view that weekly betas have become standard

practice. The UCA further explained that her comments are not reflective of a recent text which

she co-authored, “Risk and Return for Regulated Industries,” in which the use of monthly return

is cited as the common approach.467

Commission findings

340. In the 2016 GCOC proceeding, witnesses recommended beta estimates in the range of

0.45 to 0.92. In its decision, the Commission observed that all witnesses had employed methods

to estimate beta that were generally accepted, but that the resulting beta range of 470 bps was

substantially wider than the beta range of 250 bps in the 2013 GCOC proceeding. The

Commission found that it could not identify, with any reasonable degree of confidence, a method

that allowed the Commission to narrow the range of betas recommended by the witnesses.468

341. In its April 20, 2017 letter469 initiating this proceeding, the Commission stated the

following:

(i) If there is a wide range of beta values provided by the experts, will the Commission

be able to identify, with any reasonable degree of confidence, a method that allows

the Commission to narrow the range of these betas?

342. In addition to updating their proxy groups, witnesses in this proceeding also presented

evidence with respect to the use of weekly versus monthly betas, and the use of the Blume

adjustment, to address the above referenced issue.

343. With respect to the use of weekly versus monthly betas, the Commission notes a strong

preference for weekly betas by the utility witnesses, whereas Dr. Cleary maintained that there is

no clearly superior method. It is clear from the evidence on the record, just as it was in the 2016

GCOC proceeding, that weekly betas are associated with values toward the higher end of the

recommended beta range, whereas monthly betas are associated with values toward the lower

end of the recommended beta range.

463 Exhibit 22570-X0131, PDF page 51. 464 Exhibit 22570-X0131, PDF page 53. 465 Transcript, Volume 10, pages 2123-2124. 466 Exhibit 22570-X0913, PDF page 24. 467 Exhibit 22570-X0897.01, PDF page 25. 468 Decision 20622-D01-2016, PDF page 46. 469 Exhibit 22570-X0078, PDF page 1.



344. The Commission is not persuaded that weekly betas are clearly superior in all instances to

monthly betas as there remains some uncertainty and disagreement in the evidence on this point.

Indeed, practitioners continue to use both weekly and monthly data, and the investment research

firms that provide the data upon which analysts rely continue to provide both weekly and

monthly data, including betas derived from both weekly and monthly data. Accordingly, the

Commission will continue to consider both weekly and monthly based beta estimates in

determining reasonable beta estimates.

345. There was also considerable debate in this proceeding over the use of the Blume

adjustment. In the 2013 GCOC decision, the Commission acknowledged that adjusted betas are

widely disseminated to investors by investment research firms, including Bloomberg, Value Line

and Merrill Lynch. However, the Commission also indicated continued uncertainty about

whether an adjustment is warranted for the betas of regulated utilities.470

346. The Commission has not been persuaded in this proceeding that adjusted betas are

superior to unadjusted betas in the context of regulated utilities. The Commission continues to

hold the view expressed in the 2016 GCOC proceeding that both raw betas and adjusted betas

provide useful information with respect to utility risk.471

347. The low end of Dr. Cleary’s recommended beta range, 0.30, was developed based on the

long-term average over the last 25-28 years. The Commission agrees with Dr. Villadsen’s

submission that this estimate is biased downward since it considers anomalous periods such as

the aftermath of the dot-com bubble.472 Even Dr. Cleary acknowledged during the 2016 GCOC

proceeding that betas from 1998 to 2002 were not meaningful.473

348. Dr. Villadsen stated that each beta calculated for up to five years after the dot-com bubble

era is still contaminated by the anomalous data. She added that in order to eliminate the trailing

impact of this anomalous data, a credible long-term average would have to exclude betas

measured using data from 1998 to 2007.474 The Commission agrees. The Commission derived the

resulting long-term average beta for the TSX Utility sub-index, excluding data from 1998 to

2007. The resulting average was 0.47.475 Based on this, the Commission finds that the low-end

beta of 0.30 proposed by Dr. Cleary can be excluded.

349. The Commission considers Dr. Cleary’s recommended beta of 0.45 to be the lower bound

for a reasonable range of betas values. The 0.45 value does not significantly differ from the

Commission’s recalculated long-term average beta of 0.47, nor the 0.43 average Dr. Cleary

calculated for his nine company Canadian utility proxy group.476 The Commission also considers

that the value-weighted portfolio beta value associated with Dr. Villadsen’s Canadian utility

470 Decision 20622-D01-2016, pdf 46. 471 Decision 20622-D01-2016, pdf 46. 472 Exhibit 22570-X0767.01, PDF page 135. 473 Exhibit 22570-X0786.01, A10. 474 Exhibit 22570-X0786.01, A10. 475 Using data from Exhibit 20622-X0464, worksheet Rolling Results. Dr. Cleary, in Exhibit 22570-X0565, PDF

page 2, references Figure 6 from Exhibit 20622-X0457. Exhibit 20622-X0457 was Dr. Villadsen’s rebuttal

evidence from the 2016 GCOC proceeding. The working paper underlying Figure 6 of her rebuttal evidence is

in Exhibit 20622-X0464. 476 Exhibit 22570-X0562.01, Table 8.



proxy group, 0.95, represents an upper bound for a reasonable range of betas. Accordingly, the

Commission considers that a reasonable range of betas is 0.45 to 0.95.

8.2.4 The resulting CAPM estimate

350. The witnesses in this proceeding presented a large number of CAPM estimates by

varying input parameters for each of the risk-free rate, beta and MERP.

351. Table 4 below summarizes the CAPM estimates and input parameters.

Table 4. CAPM inputs and resulting ROE estimates

Rf Beta MERP Float Adj ROE

(%) (%) (%)

Cleary-recommendation 2.60 0.450 5.00 0.50 0.13% 5.48

Coyne-Canadian utility proxy group 3.26 0.749 7.70 0.50 N/A 9.53

Coyne-U.S. electric proxy group 3.85 0.666 7.70 0.50 N/A 9.49

Coyne-North American electric proxy group 3.73 0.682 7.70 0.50 N/A 9.48

Hevert-Canadian utility proxy group 3.08 0.717 9.14 0.50 N/A 10.13

Hevert-U.S. utility proxy group-2018 3.30 0.624 10.02 0.50 N/A 10.08



Villadsen-Scenario 1-Cdn. utility proxy group-low 3.45 0.850 5.70 0.50 N/A 8.80

Villadsen-Scenario 1-Cdn. utility proxy group-high 3.45 0.950 5.70 0.50 N/A 9.40

Villadsen-Scenario 1-U.S. gas LDC utility proxy group-low 3.45 0.663 5.70 0.50 N/A 7.70

Villadsen-Scenario 1-U.S. gas LDC utility proxy group-high 3.45 0.669 5.70 0.50 N/A 7.80

Villadsen-Scenario 1-U.S. electric utility proxy group-low 3.45 0.578 5.70 0.50 N/A 7.20

Villadsen-Scenario 1-U.S. electric utility proxy group-high 3.45 0.608 5.70 0.50 N/A 7.40

Villadsen-Scenario 1-U.S. water utility proxy group-low 3.45 0.644 5.70 0.50 N/A 7.60

Villadsen-Scenario 1-U.S. water utility proxy group-high 3.45 0.750 5.70 0.50 N/A 8.20

Villadsen-Scenario 2-Cdn. utility proxy group-low 3.30 0.850 8.00 0.50 N/A 10.60

Villadsen-Scenario 2-Cdn. utility proxy group-high 3.30 0.950 8.00 0.50 N/A 11.40

Villadsen-Scenario 2-U.S. gas LDC utility proxy group-low 3.30 0.663 8.00 0.50 N/A 9.10

Villadsen-Scenario 2-U.S. gas LDC utility proxy group-high 3.30 0.669 8.00 0.50 N/A 9.20

Villadsen-Scenario 2-U.S. electric utility proxy group-low 3.30 0.578 8.00 0.50 N/A 8.40

Villadsen-Scenario 2-U.S. electric utility proxy group-high 3.30 0.608 8.00 0.50 N/A 8.70

Villadsen-Scenario 2-U.S. water utility proxy group-low 3.30 0.644 8.00 0.50 N/A 9.00

Villadsen-Scenario 2-U.S. water utility proxy group-high 3.30 0.750 8.00 0.50 N/A 9.80

Commission findings

352. The results in Table 4 above, show a wide range of ROE estimates based on CAPM,

ranging from 5.48 per cent (Dr. Cleary’s recommended value) to 11.40 per cent (Villadsen-

Scenario 2-Canadian utility proxy group, which uses the value-weighted portfolio beta of 0.950).

353. The wide range of CAPM estimates is not surprising, given the Commission’s

determination above that the use of weekly and monthly based beta estimates, as well as the use

of adjusted and unadjusted betas in the CAPM model, are acceptable. This wide range of CAPM

results does not, on its own, provide much assistance to the Commission in determining an

approved ROE. Nonetheless, the Commission has determined, as discussed below, a point

estimate of 7.90 per cent with respect to the CAPM, which it will consider in establishing the

ROE fair return for the affected utilities.



354. Given the uncertainty regarding the magnitude and timing of potential changes in the

risk-free rate, as discussed in Section 8.2.1, the Commission considers the estimate of 2.60 per

cent recommended by Dr. Cleary to be reasonable. With respect to beta, as explained in

Section 8.2.3, the Commission found that the low-end beta of 0.30 proposed by Dr. Cleary can

be excluded. Further, as explained in Section 8.1, the Commission will disregard any beta

coefficients derived in connection with Dr. Villadsen’s U.S. pipeline proxy group and its

subsample group. From the remaining betas, the Commission has determined an average beta of

0.686.477 In Section 8.2.2, the Commission’s findings suggested a MERP that is in excess of

6.89 per cent. The Commission considers a MERP of 7.00 per cent to be reasonable for

determining a point estimate. Using the risk-free rate of 2.60 per cent, along with a MERP of

7.00 per cent, an average beta of 0.686 and allowing for a flotation allowance of 0.50 per cent

results in an ROE estimate of 7.90 per cent.

355. In the 2016 GCOC decision, the Commission noted that Dr. Booth placed less weight on

his CAPM models due to abnormally low interest rates.478 The Commission also noted

Dr. Villadsen’s testimony in the 2016 GCOC proceeding that she had placed less weight on her

CAPM models than in the past.479 In the current proceeding, Mr. Hevert indicated that he had

placed less weight on his CAPM model as well.480 The Commission considers that while interest

rates have risen somewhat since the time of the 2016 GCOC proceeding, they are still low

relative to average historical rates and accordingly, the Commission will give less weight to the

CAPM ROE results put forward in this proceeding.

356. The Commission also gave less weight to parties’ CAPM estimates in the 2016 GCOC

decision, compared to the CAPM estimates in the 2013 GCOC proceeding, largely due to the

Commission’s finding that it could not identify, with any reasonable degree of confidence, a

method that allowed the Commission to narrow the range of betas recommended by the experts

in that proceeding.481 Also, given that the range of betas has increased slightly in this proceeding

(even after the Commission rejected certain results on the extreme ends of the original range

presented, as discussed above), the relatively wide range of betas, compared to the 2013 GCOC

proceeding, continues to be a factor that leads the Commission to assign less weight to the

CAPM ROE results.

8.2.5 The ECAPM

357. Consistent with their evidence filed in the 2016 GCOC proceeding, Dr. Villadsen and

Mr. Hevert noted that empirical research has shown that the actual security market line (SML)

described by the CAPM formula is not as steeply sloped as the predicted SML. In other words,

low-beta securities earn returns somewhat higher than CAPM would predict, and high-beta

477 This is the average of the following betas: 0.450 recommended by Dr. Cleary; 0.749 from Mr. Coyne’s

Canadian utility proxy group; 0.666 from Mr. Coyne’s U.S. electric proxy group; 0.717 from Mr. Hevert’s

Canadian utility proxy group; 0.624 from Mr. Hevert’s U.S. utility proxy group; 0.850 from Dr. Villadsen’s

Canadian utility proxy group; 0.950 from Dr. Villadsen’s portfolio beta for her Canadian utility proxy group;

0.669 from Dr. Villadsen’s U.S. gas LDC utility proxy group; 0.663 from Dr. Villadsen’s portfolio beta for her

U.S. gas LDC utility proxy group; 0.614 from Dr. Villadsen’s U.S. electric utility proxy group; 0.578 from

Dr. Villadsen’s portfolio beta for her U.S. electric utility proxy group; 0.750 from Dr. Villadsen’s U.S. water

utility proxy group; 0.644 from Dr. Villadsen’s portfolio beta for her U.S. water utility proxy group. 478 Decision 20622-D01-2016, paragraph 311. 479 Decision 20622-D01-2016, paragraph 309. 480 Exhibit 22570-X0153.01, PDF page 95. 481 Decision 20622-D01-2016, paragraph 317.



securities earn returns somewhat lower than predicted.482 The ECAPM adds an empirical

adjustment factor to CAPM (referenced as “X” by Mr. Hevert and as “alpha” by Dr. Villadsen)

intended to adjust the SML to account for the difference between the predicted returns for a

given beta when using CAPM and future, realized returns for the same or similar beta.483

358. As in the 2016 GCOC proceeding, both Mr. Hevert and Dr. Villadsen relied on the use of

the ECAPM in developing their ROE estimates, although their models were of a different form

and used different notation. These witnesses confirmed there was no conflict between their two

approaches.484

359. In applying his version of the ECAPM, Mr. Hevert used an empirical factor of 0.25,

based on the published work of Dr. Morin. Mr. Hevert’s ECAPM results averaged 9.57 per cent

in 2017 and 10.27 per cent in 2018 for Canada, and were in the 10.00 to 11.50 per cent range for

the U.S. over the 2017-2020 period. These results were approximately 60 bps higher than his

estimates using CAPM for his Canadian utility proxy group and 100 bps higher than for his U.S

utility proxy group.

360. Dr. Villadsen used an alpha factor of 1.5 per cent, based on an average adjustment factor

from academic literature, which she further adjusted downward to account for differences in

government bond maturities and to be conservative. Dr. Villadsen’s ECAPM results were 10 to

20 bps higher than the CAPM results for her Canadian utility proxy group and were

approximately 40 to 70 bps higher than the CAPM results for her three main U.S proxy groups.485

361. Dr. Cleary stated that “Using the [ECAPM] also implicitly adjusts the beta used in

traditional CAPM estimates. Hence, the ECAPM should also not be used.”486 Mr. Hevert and

Dr. Villadsen disagreed with this conclusion.

362. In argument, Calgary pointed out that in the 2004 GCOC decision, the Commission’s

predecessor stated:

The Board notes Calgary/CAPP’s [Canadian Association of Petroleum Producers]

argument that applying CAPM using long-term interest rates (long-Canada bond yields)

in determining the risk-free rate, as was done by all experts in this Proceeding, already

corrects for the alleged under-estimation that ECAPM was designed to address.

Calgary/CAPP argued that the under estimation would only be present if the CAPM were

applied using short-term interest rates, which none of the experts did in this Proceeding.

The Board finds the Calgary/CAPP position persuasive and considers that the use of

long-term Canada bond yields largely adjusts for the tendency of CAPM, when based on

short-term interest rates, to under estimate the required returns for lower risk companies.

Therefore, the Board will only place limited weight on the results of the ECAPM

model.487

482 Exhibit 22570-X0153.01, PDF page 84. 483 Exhibit 22570-X0153.01, PDF pages 84-85. Exhibit 22570-X0193.01, PDF page 62. 484 Transcript, Volume 4, page 680. Transcript, Volume 6, page. 485 U.S. electric utility proxy group, U.S. gas local LDC utility proxy group, U.S. water utility proxy group. 486 Exhibit 22570-X0562.01, PDF page 50. 487 Decision 2004-052: Generic Cost of Capital, Application 1271597-1, July 2, 2004, page 22.



363. Calgary further submitted that “ECAPM is not academically respected, is not in

textbooks and has not been a topic in finance research … which has long since moved into

looking at multi-factor models.”488 If any improvement over the CAPM were required, Calgary

appeared to express its preference for using multi-factor models such as the Fama-French

model.489

364. The UCA raised similar points and endorsed Dr. Cleary’s view that the ECAPM “is not

very widely used in practice.”490 The UCA also pointed to the “dated nature of the literature” or

research underlying the adjustment factors used by Dr. Villadsen and Mr. Hevert. For these

reasons, the UCA recommended the Commission not place any weight on the results obtained

using the ECAPM.491

Commission findings

365. In the 2016 GCOC decision, the Commission stated:

199. In the Commission’s view, the ECAPM appears to be a model that could

contribute to the Commission’s determination of a fair allowed ROE. Generally speaking,

the Commission is supportive of models and methods that attempt to improve upon

CAPM results. The Commission agrees with Mr. Hevert that the selection of an empirical

adjustment factor is a matter of judgement. Based on the evidence in this proceeding,

however, the Commission has been unable to assess adequately the empirical adjustment

factors employed by the experts in exercising their judgement. Consequently, the

Commission will not rely heavily on the ECAPM results in this proceeding. In order for

the Commission to adequately assess the judgement exercised by the experts, the

Commission would require a full explanation justifying the sample and time periods

adopted.

366. The Commission benefitted from the evidence provided by parties on ECAPM in this

proceeding, including the information provided by Mr. Hevert492 and Dr. Villadsen493 on the

sample and time periods utilized by the studies supporting the ECAPM. In an exchange with

Commission counsel, Dr. Villadsen confirmed that the alpha factors that she relied on are all

based on studies prior to 1991, because academic studies have not studied the alpha parameter

since then.494 In an exchange with Calgary counsel, Dr. Villadsen stated:

Q. And has ECAPM ever been criticized in journals, financial journals, to your

knowledge?

A. DR. VILLADSEN: I don't think the ECAPM has been the topic of discussion in

journals that I have reviewed recently.

Q. Okay.

A. DR. VILLADSEN: Most have switched to doing multifactor models.495

488 Exhibit 22570-X0903, paragraph 54. 489 Exhibit 22570-X0903, paragraphs 52 and 55. 490 Transcript, Volume 10, page 2143. 491 Exhibit 22570-X0897.01, paragraphs 48, 150 and 153. 492 Exhibit 22570-X0159, PDF pages 19-21. 493 Exhibit 22570-X0192.01, PDF pages 28 to 30. Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-011(b),

attachment is in Exhibit 22570-X0309. 494 Transcript, Volume 4, page 680. 495 Transcript, Volume 2, page 263.



367. Recognizing that Dr. Morin’s data was from the period 1926 to 1984, Mr. Hevert

performed his own analysis using data over the 10-year period ending in 2016 to confirm that the

premise behind ECAPM is still valid. However, while he was able to confirm the general

premise of the ECAPM model, Mr. Hevert acknowledged that his analysis was not designed to

confirm the reasonableness of Dr. Morin’s alpha coefficients, which he adopted.496

368. The Commission also finds informative Mr. Hevert’s and Dr. Villadsen’s explanations

that multi-factor models aim to address the same issue as the ECAPM – specifically, to correct

for the fact that the SML is flatter than CAPM predicts, or more generally, to capture asset

pricing more accurately than CAPM.497 While ECAPM performs this correction by way of an

empirical adjustment parameter (alpha), the multi-factor models do so by employing several

parameters in addition to beta.498

369. Based on the evidence in this proceeding, it appears to the Commission that ECAPM

attempts to provide a practical solution by introducing an empirical adjustment factor to the

CAPM results; increasing the CAPM return estimates for companies with betas lower than one

and decreasing the return estimates for companies with betas higher than one. As stated by

Dr. Villadsen, “you can think of it as the ECAPM is a shortcut to multifactor models.”499

However, as the Commission pointed out in the 2016 GCOC decision, these adjustment factors

are a function of the sample and time period over which the returns were examined, as well as

the assumptions employed.500

370. Dr. Cleary questioned whether using ECAPM and adjusted betas at the same time

“essentially adjusts raw betas twice.”501 Mr. Hevert and Dr. Villadsen maintained that the use of

the ECAPM and adjusted betas are meant to address two different issues. They indicated that the

studies underlying the theoretical use of the ECAPM did not use adjusted betas to arrive at the

empirical adjustment factors.502 Consequently, different empirical adjustment factors of ECAPM

may need to be employed when applied to adjusted betas or, conversely, unadjusted betas may

need to be employed in any future ECAPM that relies on the empirical adjustment factors used

by Dr. Villadsen and Mr. Hevert.

371. The Commission further observes that all the studies on which Dr. Villadsen relied to

determine her empirical adjustment factors relied on monthly stock returns for all stocks traded

on the major U.S. stock exchanges. Given that, in this proceeding, Mr. Hevert and Dr. Villadsen

employed weekly betas, this may result in a further mismatch. It is also possible that some other

modifications to the empirical ECAPM adjustment coefficients may be required, unique to

regulated utilities, as the original ECAPM studies from which Mr. Hevert and Dr. Villadsen

obtained their adjustment factors, focused on a wide range of companies traded on the equity

market.

496 Transcript, Volume 6, pages 1217-1219. 497 Transcript, Volume 4, page 683. Transcript, Volume 6, page 1221. 498 Exhibit 22570-X0918, paragraph 153. 499 Transcript, Volume 4, page 684. 500 Decision 20622-D01-2016, paragraph 197. 501 Exhibit 22570-X0562.01, PDF page 50. 502 Transcript, Volume 4, page 682. Transcript, Volume 6, page 1221.



372. The Commission also remains of the view, expressed in paragraph 200 of the 2016

GCOC decision,503 that the empirical adjustment factor in ECAPM does not resolve the issue

regarding the wide range of estimated betas.

373. For the above reasons, the Commission will not assign significant weight to the ECAPM

results in this proceeding. The Commission acknowledges the practical difficulties associated

with using the multi-factor models described by Dr. Villadsen at the hearing, such as the need for

more data and the need to estimate not just one, but three to four parameters.504 Nevertheless, the

Commission considers it preferable to improve the CAPM results by way of multi-factor models

that specifically aim to identify factors explaining the required return, if possible, rather than

using empirical adjustment factors as is done under the ECAPM.

8.3 Other risk premium models

374. In addition to CAPM and ECAPM, parties relied on other risk premium models.

Mr. Hevert and Dr. Cleary explained that risk premium models are based on the basic financial

principle that since stocks are riskier than bonds, investors will require a higher return to invest

in a firm’s stock than in its bonds.

375. As in previous GCOC proceedings, Dr. Cleary employed a BYPRPM in developing his

ROE recommendation. Dr. Cleary explained that under his method, a risk premium in the two to

five per cent range is added to the yield on a firm’s outstanding publicly traded, long-term bonds

to arrive at a company’s cost of equity estimate, with 3.5 per cent generally added to reflect

average risk companies, and lower values added for less risky companies. Given the low-risk

nature of Canadian regulated utilities, Dr. Cleary opined that an appropriate risk premium for

these companies would be in the two- to three-per-cent range, with a best estimate of 2.5 per

cent.

376. Dr. Cleary noted that as of November 15, 2017, the yield on long-term A-rated Canadian

utility bonds was 3.51 per cent according to the Bloomberg data. Because this number was close

to the yields on outstanding Canadian utility bonds, Dr. Cleary concluded that the 3.5 per cent

bond yield was a reasonable starting point for his BYPRPM estimate. After adding his risk

premium estimate of 2.5 per cent, Dr. Cleary obtained an ROE estimate of 6.50 per cent,

inclusive of the 50 bps flotation allowance.505

377. Mr. Hevert employed the two risk premium models that he used in the 2016 GCOC

proceeding; the Predictive Risk Premium Model (PRPM) applied to his Canadian and U.S. proxy

groups, and the BYPRPM, based on approved returns for U.S. electric utility companies.

378. The PRPM, also referred to in the literature as the “general consumption-based asset

pricing model,”506 estimates the equity risk premium through the prediction of volatility.

Specifically, the risk premium derived from the PRPM is based on the premise that the volatility

of stock returns and risk premiums changes over time and is related from one period to the next.

As such, it can be estimated by using time series analysis tools such as the autoregressive

conditional heteroscedasticity (ARCH) model, and its generalized form, the GARCH model. The

503 Decision 20622-D01-2016. 504 Transcript, Volume 4, pages 683-684. 505 Exhibit 22570-X0562.01, PDF pages 66-67. 506 Transcript, Volume 6, page 1239.



inputs to the PRPM-derived model are the historical returns on the common shares of each proxy

company, less the historical monthly yield on long-term government bonds. Using statistical

software, Mr. Hevert calculated each proxy company’s projected risk premium, which he then

added to his recommended average risk-free rates.

379. Mr. Hevert also employed a variant of the BYPRPM that adds a risk premium, calculated

as the difference between approved ROEs granted by U.S. regulators and the then-prevailing

level of the long-term Treasury yield, to a long-term government bond yield.

380. Mr. Hevert modelled the relationship between interest rates and the risk premium using

regression analysis, in which the observed risk premium was the dependent variable, and the

average 30-year Treasury yield was the independent variable. According to Mr. Hevert, his

regression analysis demonstrated that over time there has been a statistically significant, negative

relationship between the 30-year Treasury yield and the risk premium.

381. Mr. Hevert noted that in previous GCOC and related decisions, the Commission

expressed concerns with relying on returns approved by other regulators in determining the fair

ROE for Alberta utilities. Mr. Hevert acknowledged that his approach to BYPRPM is based on

the premise that U.S. approved returns are a proxy for required market returns. However, based

on his practical experience, Mr. Hevert believed that approved returns are a reasonable input

because “investors consider a broad range of data, including returns authorized in other

jurisdictions, in establishing their return requirements.” In addition, Mr. Hevert stated:

… Because authorized ROEs reflect both prevailing market conditions during each rate

case and the types of market-based models proposed in GCOC proceedings in Alberta, it

is reasonable to use authorized returns to estimate the relationship between interest rates

and the Equity Risk Premium. As Dr. Morin notes:

(a)llowed risk premiums are presumably based on the results of market-based

methodologies presented to regulators in rate hearings and on the actions of

objectives unbiased investors in a competitive marketplace.136

____________ 136 Roger A. Morin, New Regulatory Finance (Public Utility Reports, Inc., 2006), at 125.507

382. Mr. Coyne introduced a similar model in his rebuttal evidence, where he looked at the

relationship between risk-free rates, approved ROEs and the implied risk premium, based on

historical allowed returns from 735 U.S. electric utility company rate cases for the period 1992

through 2017. Mr. Coyne performed a regression analysis using the implied risk premium

(calculated as a difference between approved ROEs and the then-prevailing 30-year Treasury

yields) as a dependent variable and yields on 30-year Treasury bonds as an independent variable.

383. According to Mr. Coyne, the regression results confirmed that the risk premium varies

with the level of bond yield, and generally increases as the bond yields decrease, and vice versa;

specifically, a one percentage point increase in bond yield results in a 0.55 percentage point

decrease to the implied risk premium and thus leads to a 0.45 percentage point increase in the

approved ROE. Mr. Coyne claimed that the advantage of this approach is that it allows for

examination of the actual risk premium awarded to a large group of utilities over past years,

507 Exhibit 22570-X0153.01, PDF page 71.



covering several economic cycles, and the ability to measure the inverse relationship between

risk-free rates and the risk premium recognized by regulators.508

384. Based on these regression coefficients, Mr. Coyne then estimated the required ROE using

current and expected 30-year Treasury bond yields, including the current 30-day average, a near-

term Blue Chip consensus forecast for 2018, and a Blue Chip consensus forecast for 2018-2020.

385. The experts critiqued each other’s risk premium models. Dr. Villadsen, Mr. Coyne and

Mr. Hevert pointed out that Dr. Cleary used the bond yield as of November 2017 in his analysis,

rather than a forward-looking estimate applicable to the 2018-2020 test period for this

proceeding, and thus did not take into account the expected increase in interest rates. As well,

they pointed out that Dr. Cleary’s BYPRPM approach relies on a subjective 2.5 per cent risk

premium adder that does not take into account the inverse relationship between bond yields and

risk premium.509

386. Dr. Cleary, in turn, expressed his view that Mr. Hevert and Mr. Coyne have applied the

BYPRPM incorrectly:

This is incorrect, since the BYPRP model, according to the CFA [chartered financial

analyst] literature (and numerous other textbooks), and which is commonly used in

analyst reports, adds a risk premium to the present yield on a firm’s outstanding publicly-

traded long-term bonds. It therefore estimates a market-based return based on the yield on

a company’s outstanding bonds, which is reflective of market yield spreads. It does not

use government yields, nor does it use ROEs and it certainly does not use allowed ROEs.

Furthermore, the Commission has not applied allowed ROEs in other jurisdictions in

previous decisions, including the 2013 GCOC Decision and the 2016 GCOC Decision.

[footnote omitted]510

387. To address Dr. Cleary’s point, Mr. Hevert undertook an additional analysis calculating

the equity risk premium as the difference between the approved ROEs and the prevailing yield

on the Moody’s A Utility Bond Index. Mr. Hevert indicated that the results were consistent with

his original analysis, and the choice of bond yields (government vs. utility) did not alter the

underlying inverse relationship between bond yield and risk premium. Assuming Dr. Cleary’s

A-rated public utility bond yield of 3.50 percent, Mr. Hevert’s revised method produced an ROE

of 9.69 per cent, which was within his recommended range.511

Commission findings

388. In previous GCOC decisions, the Commission accepted the BYPRPM approach as a

valid tool in estimating the cost of equity as it is simple to use and conforms to the basic

principle that investors require a higher return for assets with greater risk. The evidence in this

proceeding lends further support to this conclusion.

389. The Commission agrees with the view expressed by both Mr. Hevert and Dr. Cleary that

an advantage of this method is that it incorporates readily observable, market-determined data

508 Exhibit 22570-X0775, PDF pages 20-22 with calculations provided in Exhibit 22570-X0778. 509 Exhibit 22570-X0767.01, PDF page 71. Exhibit 22570-X0775, PDF page 26. Exhibit 22570-X0741.01,

PDF page 53. 510 Exhibit 22570-X0562.01, PDF page 67. 511 Exhibit 22570-X0741.01, PDF page 54.



such as bond returns and yields, which in turn can be deconstructed into the risk-free rate and

credit spread components.512 The Commission observes that the credit spread component needed

by utility bond investors is imbedded in the return to equity investors, along with some additional

margin. However, the remaining margin, the equity risk premium, requires estimation, and thus

requires judgment.

390. The BYPRPMs presented by Mr. Hevert, Dr. Cleary and Mr. Coyne start with observable

market-based information, specifically the yield on either utility or government long-term bonds.

Where utility bonds are used (as was done by Dr. Cleary, and Mr. Hevert in his rebuttal

evidence) the bond yield also incorporates a credit spread, which the Commission in past GCOC

decisions has accepted to be an objective measure that helps inform the Commission about

investors’ risk perceptions. However, in the Commission’s view, the BYPRPMs presented in this

proceeding falter in their application of the equity risk premium adder to the bond yield.

391. In this proceeding, Dr. Cleary recommended using the same 2.5 per cent risk premium

value that he recommended in the 2013 and 2016 GCOC proceedings.513 In the 2013 and 2016

GCOC decisions, the Commission noted the ad hoc nature of Dr. Cleary’s BYPRPM approach to

the estimation of ROE.514 As well, the Commission expressed concern with the fact that this

approach does not appear to take into account the inverse relationship between the risk premium

and interest rates, and therefore, may not apply in an environment of low interest rates.515 These

concerns were shared by the utility experts in this proceeding and in the Commission’s view,

continue to apply.

392. The BYPRPMs of Mr. Hevert and Mr. Coyne estimate the risk premium component by

comparing the approved ROEs to the long-term government bond yields in place at the time, thus

capturing the inverse relationship. However, the Commission has two concerns with

Mr. Hevert’s and Mr. Coyne’s approach. First, because their models estimate the risk premium in

excess of long-term government bond yields, i.e., the risk-free rate, they lose the advantage of

incorporating the observable market data on utilities’ credit spreads, as compared to Dr. Cleary’s

approach.

393. Second, these models use the approved ROEs of other regulators in the U.S. as proxies

for the market return. In the Commission’s view, although observable, the ROEs approved for

the U.S. utilities are not strictly market data. Accordingly, the main assumption of these models,

that the approved ROEs represent market return, does not hold, because the approved ROEs

would be heavily influenced by the ROEs awarded by other regulators.

394. While Mr. Hevert expressed his belief that “authorized ROEs reflect both prevailing

market conditions during each rate case and the types of market-based models proposed in

GCOC proceedings in Alberta,”516 the Commission observed in the 2016 GCOC decision that

this may not always be the case. Approved ROEs may be established on a different basis. For

example, they may be a result of an ROE adjustment formula or a negotiated settlement, or they

may include non-market elements such as incentive mechanisms. This led the Commission to

512 Exhibit 22570-X0153.01, PDF page 68. Transcript, Volume 10, page 2184. 513 Decision 20622-D01-2016, paragraph 226. 514 Decision 2191-D01-2015, paragraphs 260-262. Decision 20622-D01-2016, paragraph 229. 515 Decision 20622-D01-2016, paragraphs 228-229. 516 Exhibit 22570-X0153.01, PDF page 71.



conclude that it “is hard to ascertain whether further adjustments to account for aberrations of

this kind are required, without scrutinizing each regulatory decision in detail.”517

395. For all of the above reasons, the Commission did not place any weight on the results of

the BYPRPMs presented by Dr. Cleary, Mr. Hevert or Mr. Coyne.

396. Nevertheless, the Commission finds part of Dr. Cleary’s BYPRPM analysis useful.

Specifically, the Commission takes note of Dr. Cleary’s observation that yields on Bloomberg

generic long-term A-rated Canadian utility bonds (which parties agreed track the yields on

Alberta utility bonds with reasonable accuracy) have been relatively stable since the time of the

2016 GCOC proceeding. According to Dr. Cleary, this stability in the overall yield was the result

of an inverse relationship between interest rates (which increased) and credit spreads (which

narrowed) over the period leading up to this proceeding.518 Figure 7 in Section 6 and underlying

data, support this observation and show that despite periodic short-term fluctuations, the yields

on Bloomberg generic long-term A-rated Canadian utility bonds have been relatively stable with

the average yield of 3.68 in 2016, 3.65 in 2017 and 3.65 in January through March of 2018.519

As such, the Commission notes that changes in the interest rate and the utility bond credit spread

appear to have offset each other to some extent.

397. Mr. Hevert’s PRPM results suggested an ROE of 10.5 to 11.3 per cent for his Canadian

utility proxy group and 10 per cent for his U.S. utility proxy group. The Commission observes

that Mr. Hevert’s Canadian PRPM ROE estimates have increased by more than 100 bps as

compared to the 2016 GCOC proceeding. The U.S. results were relatively stable around 10 per

cent in both the 2016 GCOC and the current proceeding.520

398. In the 2016 GCOC decision, the Commission assigned little weight to Mr. Hevert’s

PRPM analysis, in part due to the lack of record development in regard to his analysis, but

expressed interest in exploring these types of models in subsequent GCOC proceedings.521 In this

proceeding, Mr. Hevert provided further explanations and support for this model.522

399. The Commission acknowledges Mr. Hevert’s statement that the PRPM approach is

relatively new,523 and no evidence was presented by Mr. Hevert to indicate whether it has been

widely accepted by other utility regulators. When asked by Commission counsel whether this

approach has been adopted in any of the regulatory proceedings that Mr. Hevert or his colleagues

have proposed it in, he responded that “it has not been explicitly rejected.”524 No evidence was

presented regarding whether any issues with the PRPM were identified by parties in other

517 Decision 20622-D01-2016, paragraph 225. 518 Transcript, Volume 10, page 2087. 519 Commission staff calculations based on data in Exhibit 22570-X0836. 520 As summarized in paragraph 209 of Decision 20622-D01-2016, in the 2016 GCOC proceeding for the Canadian

utilities proxy group, Mr. Hevert calculated the average and median risk premiums to be 7.12 per cent and

6.83 per cent, respectively. By adding these risk premiums to his recommended average risk-free rate value for

Canada of 2.59 per cent, Mr. Hevert obtained ROE estimates of 9.42 per cent and 9.71 per cent. For the U.S.

utilities proxy group, the calculated average and median risk premiums were 7.15 per cent and 7.06 per cent,

respectively. When added to Mr. Hevert’s recommended average risk-free rate value for the U.S. of 3.20 per

cent, the resulting ROE estimates were 10.35 per cent and 10.26 per cent. 521 Decision 20622-D01-2016, paragraph 222. 522 Exhibit 22570-X0153.01, PDF pages 134-136. 523 Transcript, Volume 6, page 1242. 524 Transcript, Volume 6, page 1242.



proceedings where the PRPM was put forward by Mr. Hevert or his colleagues. Parties in this

proceeding did not engage in any meaningful analysis of the PRPM, and as a result the

Commission is unable to identify whether there are any theoretical constraints associated with

using this model to develop ROE estimates for regulated utilities.

400. As parties in this proceeding did not engage in any meaningful analysis of the PRPM and

no evidence was presented that the PRPM has been vetted and accepted by other utility

regulators as a valid approach to estimate ROEs for regulated utilities, the Commission is not

prepared to assign the PRPM any weight in this proceeding.

8.4 Discounted cash flow model

401. The DCF approach estimates the cost of a company’s common equity based on the

current dividend yield of the company’s shares plus the expected future dividend growth rate.

The DCF method calculates ROE as the rate of return that equates the present value of the

estimated future stream of dividends to the current share price.

402. There are several types of DCF models and variations, including single-stage growth

models and multi-stage growth models. Single-stage, constant growth models assume that

growth in dividends will occur indefinitely at the same constant rate. Multi-stage models assume

the expected dividend growth will vary over different time periods.

403. In this proceeding, both single-stage and multi-stage DCF model estimates for utility

equities and the market as a whole were presented. Mr. Coyne, Dr. Cleary, Dr. Villadsen and

Mr. Hevert submitted DCF model estimates for utility equities in order to directly estimate the

required ROE for Alberta utilities. Dr. Cleary and Mr. Hevert submitted DCF model estimates

for the market as a whole in order to gauge the reasonableness of their Alberta utilities’ ROE

estimates. Mr. Hevert also used his market DCF estimate to calculate his MERP estimate as

noted above in Section 8.2.2.

Discounted cash flow estimates – utility proxy groups

404. Dr. Villadsen used both single-stage and multi-stage DCF models to develop ROE

estimates for her utility proxy groups.

405. In her implementation of the multi-stage DCF model, Dr. Villadsen assumed that for the

first five year period, the sample companies grow their dividends at a company-specific rate of

earnings growth and then taper off over a five year period to the long-term rate of growth. For

the initial high growth period, Dr. Villadsen used investment analyst forecasts of company-

specific growth rates sourced from Value Line and Thomson Reuters Institutional Brokers’

Estimate System (IBES), which ranged from -2.0 to 15.8 per cent. For the subsequent long-term

growth rate, Dr. Villadsen used a long-term Canadian GDP growth forecast of 3.85 per cent and

a long-run U.S. GDP growth forecast of 4.35 per cent from Consensus Forecasts.525

525 Exhibit 22570-X0193.01, PDF pages 72-73.



406. For her single-stage DCF model, Dr. Villadsen used investment analyst forecasts of

company-specific growth rates sourced from Value Line and Thomson Reuters IBES. Excluding

estimates that factor adjustments for leverage, the forecasts ranged from 8.9 to 14.8 per cent.526

407. Dr. Villadsen pointed out that one issue with the input data is that it only includes cash

dividends and does not include share repurchases as a means to distribute cash to shareholders.

To the extent that a company uses share repurchases, the input data therefore understates the cost

of equity.527

408. Because the Commission previously expressed a preference for growth rates that do not

exceed long-term GDP growth, Dr. Villadsen primarily relied on her multi-stage DCF analysis in

which she estimated a range of ROEs from 8.00 to 9.75 per cent, after considering flotation costs

and without considering financial risk.528 Relative to her DCF estimates in the 2016 GCOC

proceeding (9.00-11.50 per cent,)529 Dr. Villadsen’s estimates for this proceeding have a smaller

range and have decreased.

409. Mr. Hevert used a single-stage DCF model and a multi-stage DCF model for his

Canadian and U.S. utility proxy groups. In order to avoid any biases that may arise from

anomalous or transitory events, Mr. Hevert used average market prices over 30, 90 and 180 days

ending September 29, 2017, as inputs to his constant growth DCF model. For the growth rate

input, Mr. Hevert used security analysts’ earnings per share (EPS) growth rate forecasts.

Mr. Hevert selected the maximum high and minimum low EPS growth estimates from Value

Line, Zacks and First Call for each company in his proxy groups, to calculate a range of high and

low ROE estimates. The results were an ROE range of 10.82-12.05 per cent, and 7.56-9.42 per

cent for his Canadian and U.S. utility proxy groups, respectively, using the single-stage DCF,

exclusive of flotation cost adjustments.530 Relative to Mr. Hevert’s corresponding DCF estimates

in the 2016 GCOC proceeding,531 both ROE ranges for this proceeding have decreased.

410. Mr. Hevert explained that although the model’s form focuses on dividends, the growth

rate also represents the assumed rate of capital appreciation, and “because dividends and price

appreciation are sustained by earnings growth, the assumed growth rate should represent

investors’ expectations of growth in Earnings Per Share.”532

411. Mr. Hevert acknowledged that in Decision 20622-D01-2016, the Commission did not

accept growth rate estimates greater than the long-term GDP in the single-stage model. However,

Mr. Hevert continued to disagree with this conclusion. Mr. Hevert argued that long-term GDP

growth represents the average growth of the all sectors within the economy, and thus, should not

be a ceiling for the growth component of the single-stage model. Mr. Hevert further explained

that under the single stage model’s assumptions, the growth rate equals the rate of capital

appreciation and that, over time, capital appreciation has not been constrained by GDP growth.

To demonstrate that projected EPS growth is a valid proxy for the growth rate in the constant

526 Exhibit 22570-X0193.01, PDF pages 72-75. 527 Exhibit 22570-X0193.01, PDF page 73. 528 Exhibit 22570-X0193.01, PDF page 77. 529 Decision 20622-D01-2016, Table 10. 530 Exhibit 22570-X0153.01, Table 22 and Table 23. 531 12.49-13.88 per cent for his Canadian utility proxy group, and 8.53-10.02 per cent for his U.S. utility proxy

group. Decision 20622-D01-2016, paragraph 244. 532 Exhibit 22570-X0153.01, PDF page 51.



growth DCF model, Mr. Hevert conducted an analysis that found projected EPS as the only

statistically significant predictor variable of the variables considered.533

412. Although his position remained that analysts’ projections are a valid measure of growth

for the constant growth DCF model, in order to address the Commission’s concerns regarding

the relationship between long-term earnings growth and GDP, Mr. Hevert also employed a

multi-stage DCF model to address the limitations of the single-stage model.534 Mr. Hevert

explained that the multi-stage DCF can serve as a method to assess the reasonableness of its

inputs by referencing certain market-based metrics.535

413. To apply the multi-stage method to his Canadian proxy group, Mr. Hevert applied a

Canadian long-term growth rate of 5.02 per cent based on the real GDP growth rate of 3.15 per

cent from 1961 through 2016, and an inflation rate of 1.82 percent. To apply the multi-stage

method to his U.S. proxy group, Mr. Hevert calculated a long-term growth rate of 5.35 per cent

in a manner similar to his Canadian estimate. Due to a lack of Value Line reports for companies

in his Canadian proxy group, Mr. Hevert relied on current payout ratios for the first period, and

the interpolated payout ratio for the second period. He then assumed that by the end of the

second period (i.e., the end of year 5-10), the payout ratio would converge to each group’s long-

term average.536 Applying this method, Mr. Hevert calculated results ranging from 9.77 to

10.15 per cent and 8.59 to 9.12 per cent for his Canadian and U.S. utility proxy groups,

respectively.537

414. Mr. Hevert explained that an alternative to calculating the terminal value based on

assumed GDP growth rates is to adopt one of the fundamental assumptions underlying the

constant growth DCF model, that the current P/E ratio remains constant in perpetuity. Because

the multi-stage model projects EPS in the terminal year, Mr. Hevert applied the current P/E ratio

to the projected EPS estimate to arrive at the terminal price.538 Applying this method, the results

calculated by Mr. Hevert ranged from 9.88 to 10.76 per cent and 9.69 to 11.08 per cent for his

Canadian and U.S. utility proxy groups, respectively.539 Mr. Hevert recommended that principal

weight be given to his Canadian utility proxy group DCF results that exclude sustainable

growth.540

415. Mr. Coyne used a single-stage DCF model and a multi-stage DCF model for his three

proxy groups. For his DCF analysis, Mr. Coyne calculated dividend yields for each company in

his Canadian utility proxy group and for each company in his U.S. electric proxy group, by

dividing the current annualized dividend by the average of the stock prices for the 90-trading-day

period ending August 31, 2017. Mr. Coyne calculated the constant growth DCF model estimates

using security analysts’ EPS growth rate forecasts as the growth component from SNL Financial,

533 Exhibit 22570-X0153.01, PDF page 57. Mr. Hevert conducted four separate regressions, with P/E as the

dependent variable and projected EPS, dividends per share, book value per share and the sustainable growth,

respectively, as the explanatory variables. Upon reviewing the results, Mr. Hevert found that the only

statistically significant growth rate was projected EPS. 534 Exhibit 22570-X0153.01, PDF pages 62-63. 535 Exhibit 22570-X0153.01, PDF page 62. 536 Exhibit 22570-X0153.01, PDF pages 64-65. 537 Exhibit 22570-X0153.01, PDF page 65. 538 Exhibit 22570-X0153.01, PDF pages 65-66. 539 Exhibit 22570-X0153.01, PDF page 67. 540 Exhibit 22570-X0153.01, PDF pages 6 and 56.



Value Line, Zacks and First Call for each company in the two proxy groups.541 Similar to

Mr. Hevert, Mr. Coyne explained that analysts’ earnings growth estimates are typically relied on

when using the DCF model.542

416. In order to address the limiting assumptions present in the single-stage DCF model,

Mr. Coyne developed a multi-stage model to estimate ROE. In his multi-stage model, Mr. Coyne

transitioned from near-term growth (i.e., the average of Value Line, Zacks, SNL Financial and

First Call forecasts used in the constant growth model) for the first stage of the analysis

(years 1-5), to the long-term forecast of GDP growth for the third stage of the analysis (years 11

and beyond). In his second stage, Mr. Coyne connected the near-term growth with the long-term

growth for the transitional period by changing the growth rate each year on a pro rata basis. In

the terminal stage, the dividend cash flow then grows at the same rate as GDP to perpetuity (or a

total of 200 years in the model).543

417. Mr. Coyne explained that his DCF analyses across all methods indicate an average cost

of common equity of 10.24 per cent, 8.89 per cent and 9.13 per cent for his Canadian utility

proxy group, U.S. electric proxy group and North American electric proxy groups, respectively,

inclusive of a 50 bps adjustment for flotation costs.544

418. In formulating his utility single-stage DCF model estimates, Dr. Cleary derived two

sustainable growth rates for each of the companies in his three proxy groups.545 Dr. Cleary

calculated results for all three of his proxy groups, using both sustainable growth rates. The

resulting ROEs, excluding flotation allowance, ranged from an average of 5.01 per cent for his

four company Canadian utility proxy group, to 7.24 per cent, which was the median for his seven

company Canadian utility proxy group.546 Using the mid-point of the average ROEs and the

median ROEs for his three proxy groups, Dr. Cleary determined a best-estimate single-stage

ROE of 5.90 per cent, excluding flotation allowance.547

419. Dr. Cleary also applied the multi-stage model to his three utility proxy groups. He

estimated the short-term growth rate using payout ratios and ROEs from 2016. Dr. Cleary

estimated the long-term growth rate using long-term averages for payout ratios and ROEs.548 The

resulting ROEs, excluding flotation allowance, ranged from an average of 6.30 per cent for his

nine company Canadian utility proxy group, to 7.65 per cent for his seven company Canadian

utility proxy group.549 Dr. Cleary reported his best estimate multi-stage ROE was 6.90 per cent,

excluding flotation allowance.550

420. Dr. Cleary weighted the best estimates of his single-stage DCF model, 5.90 per cent, and

his multi-stage DCF model, 6.90 per cent, equally, to arrive at an ROE of 6.40 per cent. He

added a 50 bps flotation allowance to arrive at his DCF based ROE recommendation of 6.90 per

541 Exhibit 22570-X0131, PDF page 66. 542 Exhibit 22570-X0131, PDF pages 62-63 543 Exhibit 22570-X0131, PDF page 67. 544 Exhibit 22570-X0131, PDF page 69. 545 Exhibit 22570-X0562.01, PDF pages 56-58. 546 Exhibit 22570-X0562.01, PDF page 57. 547 Exhibit 22570-X0562.01, PDF page 58. 548 Exhibit 22570-X0562.01, PDF page 58. 549 Exhibit 22570-X0562.01, Table 14. 550 Exhibit 22570-X0562.01, Table 15.



cent.551 Dr. Cleary’s DCF based ROE recommendation in the 2016 GCOC proceeding was 8.04

per cent, inclusive of a 50 bps flotation allowance.552

421. Mr. Hevert, Dr. Villadsen, Mr. Coyne and Dr. Cleary all exchanged critiques regarding

the specific DCF models employed, the inputs used and the corresponding results.

422. In his evidence, Dr. Cleary disagreed with the utilities’ experts’ use of analyst earnings

growth estimates because they were simply too high. He pointed out that his views were shared

by the Commission in the last two GCOC decisions. Dr. Cleary highlighted that in the 2016

GCOC decision, the Commission explicitly stated that it did not accept a single-stage DCF

model which uses a growth rate exceeding the long-term GDP growth rate of the economy.

Dr. Cleary submitted that Dr. Villadsen’s, Mr. Hevert’s and Mr. Coyne’s single-stage models

should be rejected in this proceeding as they all violate the aforementioned condition.553

423. Dr. Cleary explained that a similar issue arises within Dr. Villadsen’s, Mr. Hevert’s and

Mr. Coyne’s multi-stage DCF estimates. Dr. Cleary pointed out that the implied constant

perpetual growth rates used by Dr. Villadsen, Mr. Hevert and Mr. Coyne in their multi-stage

DCF models exceed estimates for Canadian nominal GDP growth.554

424. In response to this criticism, Dr. Villadsen explained that there is no reason to believe

that any one company cannot grow at a higher or lower rate than the economy in the near term.

Dr. Villadsen also pointed out that the economy of Alberta is a relevant benchmark and is

expected to grow faster than the overall Canadian GDP in the near future.555 Mr. Coyne pointed

out that in the 2016 GCOC decision, the Commission stated it would accept growth rates above

the nominal long-term GDP growth in the initial stages of the multi-stage model.556 Mr. Hevert

responded to Dr. Cleary’s criticisms by pointing out that the single-stage and multi-stage DCF

models serve separate purposes and are not meant to be equivalent.557

425. According to Dr. Villadsen, the DCF estimates put forward by Dr. Cleary were flawed

because they failed to consider the impact of share buybacks and, therefore, underestimated the

expected market returns. Dr. Villadsen expressed her disagreement with the use of the historic

average Canadian GDP growth rate as a long-term growth rate because it is a backward-looking

metric that is conceptually flawed and inconsistent with the underlying principles of the model.558

426. Regarding Dr. Cleary’s multi-stage estimates, Dr. Villadsen took issue with the use of

Canadian GDP growth in 2023-2027 from Consensus Forecasts as the short-term growth rate

and the use of historical GDP growth over an arbitrarily chosen period for the long-term growth

input, without justification as to why these inputs were used in the model.559 Dr. Villadsen also

pointed out that while Dr. Cleary criticized Consensus Forecasts for predicting government

551 Exhibit 22570-X0562.01, PDF page 60. 552 Decision 20622-D01-2016, paragraph 256. 553 Exhibit 22570-X0562.01, PDF pages 62-63. 554 Exhibit 22570-X0562.01, PDF page 64. 555 Exhibit 22570-X0562.01, PDF page 63. 556 Exhibit 22570-X0562.01, PDF page 42. 557 Exhibit 22570-X0890.01, PDF page 73. 558 Exhibit 22570-X0767.01, PDF page 63. 559 Exhibit 22570-X0767.01, PDF page 57.



yields above what occurred since the last GCOC, Dr. Cleary used their estimates for future GDP

growth.560

427. Mr. Coyne critiqued Dr. Cleary’s dismissal of analyst growth rates and his derivation of

the sustainable growth rate in his multi-stage model. Mr. Coyne pointed out that Dr. Cleary’s

derivation of the sustainable growth rate is incomplete and understates the applicable growth

rate. Mr. Coyne explained that Dr. Cleary’s derivation assumes that utilities will not issue new

financing to support growth, and that in order to properly calculate the sustainable growth rate,

long-term expected stock financing should be factored into the equation. Mr. Coyne also pointed

out that an additional issue with this formulation is that it requires ROE as an input, creating a

circularity problem.561 Mr. Hevert pointed out that Dr. Cleary’s recommended growth rates are

unreasonable and, since they are below the Bank of Canada’s target inflation of two per cent, are

negative in terms of real growth.562

428. The UCA highlighted that Dr. Cleary did not dispute that some of his growth rates would

imply a negative real rate of return; however, he stressed that other factors, such as stable

dividends, might attract investors.563

429. In response to Dr. Cleary’s view that regulated utilities should be expected to grow at a

slower pace than the overall GDP, Mr. Coyne developed a comparison of actual earnings and

dividends per share growth rates for his three proxy groups and highlighted that both earnings

and dividend growth exceeded GDP growth by a wide margin during the period analyzed.564

430. Consistent with the other utility witnesses, Mr. Hevert explained that Dr. Cleary’s DCF

estimates are understated largely because of his reliance on sustainable growth rate estimates.

Mr. Hevert explained that contrary to the premise of sustainable growth, which Dr. Cleary

applied, empirical research has demonstrated that higher growth is associated with higher payout

ratios.565

431. A summary of the DCF models ROE results are included in Table 5 below.

560 Exhibit 22570-X0767.01, PDF pages 57-58. 561 Exhibit 22570-X0562.01, PDF page 36. 562 Exhibit 22570-X0890.01, PDF page 75. 563 Exhibit 22570-X0897.01, PDF page 47. 564 Exhibit 22570-X0775, PDF pages 44-45. 565 Exhibit 22570-X0741.01, PDF pages 51-52.



Table 5. ROE results determined using various DCF models, including flotation allowance

ROE 2018 GCOC ROE 2016 GCOC566

(%)

Dr. Villadsen-recommendation-without leverage567 8.00-9.75 9.00-11.50

Mr. Hevert-Canadian utility proxy group-single-stage568 11.32-12.55 12.99-14.38

Mr. Hevert-Canadian utility proxy group-multi-stage569 10.27-10.65 N/A

Mr. Hevert-U.S. utility proxy group-single-stage570 8.06-9.92 9.03-10.52

Mr. Hevert-U.S. utility proxy group-multi-stage571 9.09-9.62 N/A

Mr. Coyne-Canadian utility proxy group-single-stage572 10.85 N/A

Mr. Coyne-Canadian utility proxy group-multi-stage573 9.63 N/A

Mr. Coyne-U.S electric proxy group-single-stage574 9.11 N/A

Mr. Coyne-U.S electric proxy group-multi-stage575 8.66 N/A

Mr. Coyne-North American electric proxy group-single-stage576 9.47 N/A

Mr. Coyne-North American electric proxy group-multi-stage577 8.79 N/A

Dr. Cleary-recommendation578 6.90 8.04

Discounted cash flow estimates – Canadian and U.S. equity markets

432. Dr. Cleary, Mr. Coyne and Mr. Hevert each provided single-stage DCF ROE estimates

for the overall equity market. Dr. Cleary also utilized a multi-stage DCF model for this purpose.

Dr. Cleary and Mr. Hevert used the results to gauge the reasonableness of their ROE estimates.

Mr. Hevert and Mr. Coyne used their results to calculate their MERP estimates.

433. Dr. Cleary equally weighted the results of his single-stage and multi-stage results, and

provided a best estimate for the Canadian market required ROE of 7.70 per cent.579 This is lower

than the 8.75 per cent figure he presented in the 2016 GCOC decision.580

434. Mr. Hevert calculated estimated total returns of 14.84 per cent and 13.83 per cent for the

S&P/TSX and S&P 500, respectively.581 In the 2016 GCOC decision, the results presented by

Mr. Hevert were 12.65 per cent and 13.78 per cent for the S&P/TSX and S&P 500,

respectively.582

566 Decision 20622-D01-2016, Table 10. 567 Exhibit 22570-X0767.01, PDF page 77. 568 Exhibit 22570-X0153.01, Table 22. 569 Exhibit 22570-X0153.01, Table 22. 570 Exhibit 22570-X0153.01, Table 23. 571 Exhibit 22570-X0153.01, Table 23. 572 Exhibit 22570-X0131, Table 15. 573 Exhibit 22570-X0131, Table 15. 574 Exhibit 22570-X0131, Table 15. 575 Exhibit 22570-X0131, Table 15. 576 Exhibit 22570-X0131, Table 15. 577 Exhibit 22570-X0131, Table 15. 578 Exhibit 22570-X0562.01, PDF page 61. 579 Exhibit 22570-X0562.01, PDF page 54. 580 Decision 20622-D01-2016, paragraph 252. 581 Exhibit 22570-X0153.01, PDF page 66. 582 Decision 20622-D01-2016, paragraph 243.



435. Mr. Coyne calculated an estimated total return of 12.64 per cent and 12.74 per cent for

the S&P/TSX and S&P 500, respectively.583

436. Dr. Cleary was the only expert to use a multi-stage model to estimate the market return.

Mr. Hevert critiqued Dr. Cleary’s estimates as too conservative, pointing out that sustainable

growth is an inferior measure of expected growth.584

Commission findings

437. The Commission was presented with ROE estimates determined using both single-stage

and multi-stage DCF models.

438. With respect to the single-stage DCF model estimates presented by Dr. Villadsen,

Mr. Coyne and Mr. Hevert, the growth rates used by each of these three witnesses in their single-

stage DCF models are in excess of the long-term GDP growth estimates they put forward.585

Consistent with its determinations in prior GCOC decisions, the Commission will not accept, in a

single-stage DCF model, the use of long-term or terminal growth rates that exceed estimates of

the nominal long-term GDP growth rate for the economy. The Commission recognizes that the

utilities are, as Dr. Cleary stated in his evidence, essentially monopolies in mature markets and,

because of this, the use of long-term growth in excess of the long-term growth of GDP is

unreasonable.586

439. With regard to the single-stage DCF model results submitted by Dr. Cleary, the

Commission notes that the implied overall average long-term growth rate across his 12 scenarios

was 1.89 per cent.587 The Commission notes that this growth rate is within the Bank of Canada’s

targeted range of one to three per cent for inflation. If long-term inflation exceeds Dr. Cleary’s

1.89 per cent long-term growth rate, this results in negative real growth. The Commission

considers that over the long term, investors would not accept the risks of equity ownership if the

expected long-term outlook for real growth was at or near negative levels. Consequently, the

Commission will not accept the single-stage DCF model results submitted by Dr. Cleary.

440. With regard to the multi-stage DCF ROE estimates submitted by Dr. Cleary,

Dr. Villadsen, Mr. Coyne and Mr. Hevert, there was disagreement among the witnesses

regarding whether it is acceptable to use growth rates above the nominal long-term GDP growth

rate, in the initial stages of a multi-stage DCF model. In the 2016 GCOC decision, the

Commission accepted that in some circumstances, the use of growth rates above the nominal

long-term GDP growth rate may be reasonable in the initial stages.588

441. In this proceeding, Dr. Villadsen contended that there is no reason to believe that any one

company cannot grow at a higher rate than the economy in the near term. She noted that

Alberta’s economy is expected to grow faster than the Canadian GDP in the near future.589 The

Commission agrees with these submissions of Dr. Villadsen, and therefore, it will accept the use

583 Exhibit 22570-X0132, worksheets JMC-3 Canada MRP and JMC-4 US MRP. 584 Exhibit 22570-X0741.01, PDF page 51. 585 Exhibit 22570-X0562.01, Table 16. 586 Exhibit 22570-X0562.01, PDF page 63. 587 Exhibit 22570-X0562.01, Table 13, average of 1.92 per cent and 1.86 per cent. 588 Decision 20622-D01-2016, paragraph 287. 589 Exhibit 22570-X0767.01, A78.



of growth rates above the nominal long-term GDP growth rate, in the initial stages of the multi-

stage DCF models used by Dr. Villadsen, Mr. Coyne and Mr. Hevert.

442. On the subject of the long-term growth rates used in the multi-stage DCF models,

Mr. Hevert used 5.02 per cent for Canada and 5.35 per cent for the U.S. He developed these

estimates using real historical GDP growth rates.590 Dr. Villadsen and Mr. Coyne used long-term

growth rates of 3.85 per cent and 3.84 per cent, respectively, for Canada.591 For the U.S.,

Dr. Villadsen and Mr. Coyne used a long-term growth rate of 4.35 per cent.592 Dr. Villadsen’s

and Mr. Coyne’s long-term growth rates were developed using information from Consensus

Forecasts.593

443. The Commission notes that Mr. Hevert’s long-term estimates are grounded in historical

data, whereas the Consensus Forecast long-term growth rate forecasts are forward looking. The

Commission finds that historical growth rates, developed over a 55-year period,594 might not be

a valid indicator or a fair representation of future growth. Consequently, the Commission prefers

the multi-stage DCF models of Dr. Villadsen and Mr. Coyne because these use forward-looking,

long-term growth estimates.

444. Mr. Hevert criticized the growth rates that Dr. Cleary used in his multi-stage DCF model.

Mr. Hevert described them as being “unduly low.”595 The Commission notes that the implied

overall average long-term growth rate across Dr. Cleary’s six scenarios was 2.83 per cent.596 The

Commission considers that this long-term growth rate is within the Bank of Canada’s targeted

range of one to three per cent for inflation. However, as noted above, if long-term inflation

exceeds Dr. Cleary’s 2.83 per cent long-term growth rate, this results in negative real growth.

Again, the Commission considers that over the long term, investors would not accept the risks of

equity ownership if the expected long-term outlook for real growth was at or near negative

levels. Consequently, the Commission will not accept the multi-stage DCF model results

submitted by Dr. Cleary.

445. The Commission finds that both Mr. Coyne’s and Mr. Hevert’s estimates of the expected

Canadian and U.S. market returns using the DCF model, which range from 12.65 to 14.84 per

cent, are too high. These results are driven by unreasonable growth rate estimates. The

Commission observes that the basis of Mr. Coyne’s estimate of the Canadian market return

relied on a sample with approximately 14 per cent of the companies having growth rates that

exceeded 20 per cent.597 Turning to Mr. Hevert’s estimate of the Canadian market return,

approximately 16.5 per cent of the companies in his sample had growth rates that exceeded

20 per cent.598 Considering that the single-stage DCF model assumes a growth rate into

perpetuity, the Commission finds the resulting estimate unrealistic, and affords Mr. Hevert’s and

Mr. Coyne’s equity market DCF estimates no weight. In addition, the Commission notes that the

expected market return rates used by Mr. Coyne and Mr. Hevert use analyst estimates of growth

590 Exhibit 22570-X0153.01, PDF page 64. 591 Exhibit 22570-X0193.01, A68. Exhibit 22570-X0131, Table 14. 592 Exhibit 22570-X0193.01, A68. Exhibit 22570-X0131, Table 14. 593 Exhibit 22570-X0193.01, A68. Exhibit 22570-X0131, PDF page 68. 594 Exhibit 22570-X0153.01, PDF page 64. 595 Exhibit 22570-X0741.01, PDF page 30. 596 Exhibit 22570-X0562.01, Table 14, using the average of 2.42 per cent, 3.57 per cent, and 2.51 per cent. 597 Exhibit 22570-X0132, Sheet JMC-3 Canada MRP. 598 Exhibit 22570-X0154.01, Sheet Sch 6 MRP TSX.



rates that far exceed GDP growth. Accordingly, the Commission finds that the expected market

return rates put forward by Mr. Coyne and Mr. Hevert are too high. No meaningful evidence was

provided that would enable the Commission to quantify the extent of the over-estimation in order

to develop a more reasonable estimate.

446. Given the foregoing, the Commission finds that the resulting range of ROE estimates is

8.00 to 9.75 per cent, which is the recommended range of Dr. Villadsen, consisting of the

8.00 per cent ROE estimate for her U.S. gas LDC utility proxy group, and the 9.75 per cent ROE

estimate for her Canadian utility proxy group, to be reasonable. However for reasons discussed

further below, the Commission finds Dr. Villadsen’s recommended range to be biased upward.

The Commission also notes that Mr. Coyne’s three multi-stage estimates all fall within this

range.

447. The 9.75 per cent upper end of Dr. Villadsen’s ROE estimate from her multi-stage DCF

model is based on the results from her Canadian utility proxy group, which consists of nine

companies. The growth rate used in the initial stage of her multi-stage DCF model for three of

the companies in her Canadian utility proxy group is in excess of 14.00 per cent, while the initial

growth rates for the other six companies range from 2.60 to 7.48 per cent, and average 4.98 per

cent. Accordingly, the Commission considers that the results of Dr. Villadsen’s multi-stage DCF

model for her Canadian utility proxy group are skewed upward because of the use of growth

rates that exceed 14.00 per cent, in combination with the small number of companies included in

this proxy group.

448. The 9.63 per cent ROE estimate from Mr. Coyne’s multi-stage DCF model for his

Canadian utility proxy group suffers from the same issue as Dr. Villadsen’s result of 9.75 per

cent. Mr. Coyne’s Canadian utility proxy group consists of five companies, and includes an

average initial growth rate estimate of 5.96 per cent. However, there are two companies in this

proxy group that have initial growth rates in excess 8.25 per cent, which is over 38 per cent

above the average of 5.96 per cent. The Commission considers that including these two

companies in such a small proxy group significantly skews the results upward.

449. The Commission considers that the 8.79 per cent ROE estimate from Mr. Coyne’s multi-

stage DCF model for his North American electric proxy group, which excludes one of the

Canadian companies that had a growth rate in excess of 8.25 per cent, and includes all of the

companies from his U.S. electric proxy group, largely mitigates the issue regarding the outcomes

associated with the use of higher than average initial growth rates and small proxy group sizes.

Accordingly, the Commission finds that 8.79 per cent is a reasonable point estimate for the

multi-stage DCF method.

8.5 Stock market return expectations of finance professionals

450. The Commission has indicated in previous GCOC decisions that it will consider evidence

regarding return expectations of market professionals in determining a fair ROE for the regulated

utilities in Alberta. However, the Commission gave little weight to this information in Decision

20622-D01-2016 because the referenced reports and articles, with one exception, predated the



2016 GCOC proceeding, making it unclear whether the expressed expectations reflected the

current expectations of market professionals at the time of that proceeding.599

451. Consistent with his evidence from past GCOC proceedings, Dr. Cleary recommended

consideration of the return expectations of market professionals in determining the fair ROE.

He stated that beliefs of professionals participating in the markets and influencing market

activity are far more relevant than market expectations developed by utilities’ experts. To this

end, Dr. Cleary provided data showing historical long-term real returns for Canadian equity

markets. This data suggested average real returns of 6.55 per cent for Canada, with a range of

estimates from 5.6 to 7.4 per cent over approximately the last 100 years. Combining these figures

with expected inflation of two per cent would suggest expected nominal returns of 8.55 per cent,

with a range of estimates from 7.6 to 9.4 per cent based solely on long-term historical results.

452. Dr. Cleary also provided 2017 publications from several sources600 that expressed an

expectation of long-term market returns for Canada in the nominal range of 4.0 to 8.1 per cent,

with an average of 5.83 per cent. In response to a Commission IR, Dr. Cleary provided updated

numbers for some of these reports; however, he confirmed that the updated documents do not

alter his conclusions.601 Dr. Cleary then subtracted an expected inflation rate of 2.0 per cent to

arrive at an average real return of 3.83 per cent, which he pointed out was below the long-term

average for the Canadian market:

Deducting the 2% expected inflation, this translates to an average real return of 3.83%. In

other words, most market professionals are of the belief that Canadian stocks are unlikely

to earn their historic long-term real rates of return in the 5.6-7.4% range over the next

5-10 years, with most of them citing the current low interest rate environment as one of

the main contributing factors.602

453. Dr. Cleary expressed his view that both historical returns and current expectations of

market professionals represent the best sources of information regarding future long-term market

returns. Combining the historical returns and market forecasts for Canada, Dr. Cleary arrived at a

market return range of 6-9 per cent, with a midpoint of 7.5 per cent.603

454. In reference to the return expectations by market professionals provided by Dr. Cleary,

Mr. Coyne stated that such data is conservative and targeted for pension fund managers, and

while he saw no reason not to consider this information, he submitted that he would not place

primary reliance on this data.604

455. Mr. Hevert did not share the view that return expectations of market professionals should

be considered when determining a fair ROE for the Alberta utilities. In addition to pointing out

that the Commission in previous GCOC decisions indicated that pension fund managers tend to

be somewhat conservative, Mr. Hevert noted that fund managers must consider a measure of

expected returns, whereas the cost of equity is a measure of investors’ required returns.

599 Decision 20622-D01-2016, paragraph 297. 600 The Financial Planning Standards Council; consulting firms such as AON Hewitt and McKinsey; and several

investment management firms such as CIBC Asset Management, BlackRock, etc. 601 Exhibit 22570-X0675, UCA-AUC-2018JAN26-004, PDF page 9. 602 Exhibit 22570-X0562.01, PDF page 36. 603 Exhibit 22570-X0562.01, PDF page 36. 604 Transcript, Volume 5, pages 976- 977.



A pension fund asset manager will match the expected returns available from various

asset classes to the expected liabilities that must be funded. An investor seeking to

maximize his risk-adjusted return will only invest in a security if the expected return is

equal to or greater than the required return. If it is not, the investor will look to alternative

investments for which the expected return is compensatory relative to the expected risks.

Because expected returns may or may not equal required returns, it is not clear that

pension funding assumptions (that is, expected returns) should be viewed as a measure of

investors’ required returns.605

456. To understand whether the use of market expected returns is an approach endorsed by the

finance industry, Mr. Hevert conducted a review of articles published in financial journals, as

well as various texts. Mr. Hevert’s review showed that analyses of expected market returns, or

pension fund assumptions, were not among the analytical techniques used by the authors in the

determination of the cost of capital.606 Mr. Thygesen questioned Mr. Hevert’s conclusion by

pointing out that the mere absence of applying market expected returns by the authors does not

mean that the approach is irrelevant.607

457. Mr. Hevert pointed out that several of the documents relied upon by Dr. Cleary “contain

clearly stated limiting assumptions and disclaimers, which call into question their use for the

purpose of setting the ROE in this proceeding.”608 Mr. Hevert also expressed the view that in

establishing their return requirements, investors use the growth rate projections by analysts that

cover the individual stocks rather than broad market projections like those provided by

Dr. Cleary. Mr. Hevert concluded his rebuttal evidence on this subject with a reference to 2017

Duke Chief Financial Officer survey results projecting average and median hurdle rates of

17.44 per cent and 15.00 per cent, respectively, in Canada, and 13.50 per cent and 12.0 per cent

in the U.S.609

458. In a similar vein, Mr. Buttke submitted that long-term market aggregate expectations

should not be assumed to be similar to investors’ required equity returns. In Mr. Buttke’s view,

this relationship presumes that investors (including pension funds) passively invest in a given

country’s market indices, are willing to accept the public market’s aggregate return over the long

term and do not make any dynamic market decisions. Accordingly, it is inferred that Canadian

equity market investors are unable to purchase equities from other comparable markets with

higher long-term expected returns. Given the global nature of capital markets, Mr. Buttke

concluded that it is not supportable to assume that a local expected return would define

investors’ hurdle rates.610

459. At the hearing, Dr. Villadsen added that it is very difficult to sample all relevant

information and it can be challenging to figure out what is representative of the market.611

605 Exhibit 22570-X0153.01, PDF page 88. 606 Exhibit 22570-X0153.01, PDF pages 88-90. 607 Exhibit 22570-X0551, paragraph 206. 608 Exhibit 22570-X0741.01, PDF page 44. 609 Exhibit 22570-X0741.01, PDF page 25. 610 Exhibit 22570-X0749, PDF pages 98-99. 611 Transcript, Volume 4, page 652.



Commission findings

460. Consistent with its determinations in previous GCOC decisions, the Commission

continues to hold the view that return expectations of finance market professionals are germane

to the determination of a fair ROE for regulated utilities, while keeping in mind the purpose and

limitations of such estimates.

461. Regarding Mr. Hevert’s point that fund managers must consider a measure of expected

returns, whereas the cost of equity is a measure of investors’ required returns, the Commission

has discussed in Section 4 of this decision that one of the elements of the fair return standard is

investments with comparable risk. The Commission considers that the expected returns for the

equity market as a whole provide a useful reasonableness check for the fair return established for

the affected utilities. In Decision 2004-052, the board determined that it is reasonable to “expect

the required return for utilities to be below the required overall equity market return,”612 given

that investments in utility stocks are typically less risky than investments in the average company

stock in the market. The Commission agrees.

462. The Commission also acknowledges Mr. Buttke’s view with respect to the relationship

between capital markets and the option for investors to seek alternatives in markets with higher

expected returns than available in Canada. In previous GCOC decisions, the Commission

communicated that in determining a fair return for Alberta utilities, it is reasonable to rely on the

U.S. market return data given the globalization of the world economy and the integration of

North American capital markets.613

463. Notwithstanding that the market return expectations of finance professionals may be of

some informational value in the determination of a fair ROE for regulated utilities, in this

proceeding the evidence received was of little assistance to the Commission for the following

reasons.

464. Mr. Hevert provided data showing hurdle rates in the range of some 15 to 17 per cent in

Canada, and 12 to 13 per cent in the U.S. In Decision 2191-D01-2015, the Commission agreed

with those parties who stated that caution needs to be exercised when comparing hurdle rates to

the cost of equity estimates. This is because hurdle rates are often project-specific, whereas the

objective of the ROE estimation models (and a GCOC proceeding in general) is to estimate the

cost of capital for the company as a whole.614 Further, the Commission finds the results of

Mr. Hevert’s review of financial journals do not lead to the conclusion that expected market

returns and pension fund assumptions are not relevant in the determination of cost of capital.

465. In the 2016 GCOC decision, the Commission expressed concerns with the potential

suitability of the reports cited by the intervener witnesses because only one report was published

since the time of the 2013 GCOC decision. In the current proceeding, Dr. Cleary addressed this

concern and presented reports published in 2017. The Commission takes note of Dr. Cleary’s

statement that “most market professionals are of the belief that Canadian stocks are unlikely to

earn their historic long-term rates of return in the 5.6-7.4% range over the next 5-10 years, with

most of them citing the current low interest rate environment as one of the main contributing

612 Decision 2004-052, page 29. 613 Decision 20622-D01-2016, paragraph 302 with reference to Decision 2009-16, paragraph 200. 614 Decision 2191-D01-2015, paragraph 69.



factors.”615 Dr. Cleary used this information, along with the historical market returns, to arrive at

his point estimate of 7.5 per cent return for the Canadian market.

466. The Commission finds that Dr. Cleary’s point estimate of 7.50 per cent for the expected

Canadian market return is too low. Subtracting Dr. Cleary’s risk-free rate recommendation of

2.60 per cent from his point estimate of 7.50 per cent, results in a MERP of 4.90 per cent. The

Commission finds that this is much lower than the suggested minimum MERP value of 6.89 per

cent, as discussed in Section 8.2.2, and the Commission’s MERP value of seven per cent it used

in determining its CAPM point estimate. Accordingly, the Commission will not place any weight

on Dr. Cleary’s point estimate of 7.50 per cent for the expected Canadian market return, in

determining the approved ROE.

8.6 Flotation allowance

467. ROE estimates obtained through CAPM, DCF or risk premium models are usually

adjusted upward by a “flotation allowance” or “flotation costs.” The Commission noted in

previous GCOC decisions that a flotation allowance is normally included in the allowed return to

account for administrative costs and equity issuance costs, any impact of under-pricing a new

issue, and the potential for dilution.616 In the 2016 GCOC decision, the Commission found that a

flotation allowance of 50 bps was reasonable and consistent with the historical practice of the

Commission and its predecessor.

468. In this proceeding, Dr. Villadsen,617 Mr. Hevert,618 Mr. Coyne619 and Dr. Cleary620 adopted

the 0.50 per cent flotation costs adjustment allowed by the Commission in previous GCOC

decisions, including the most recent Decision 20622-D01-2016.621

Commission findings

469. The Commission finds that a flotation allowance of 0.50 per cent continues to be

reasonable and will accept this adjustment to the ROE results obtained through CAPM, DCF or

risk premium models.

8.7 Other considerations in establishing a fair approved return on equity

470. In addition to the models and information discussed in previous sections of this decision,

parties employed other considerations in arriving at their recommendations regarding a fair ROE

for the Alberta utilities.

471. Dr. Villadsen indicated that because investors compare returns across jurisdictions, it is

important to recognize the ROE and capital structures that utilities have recently been granted in

other jurisdictions.622 Therefore, she presented information on the approved ROE and capital

structure for other Canadian and U.S. utilities for 2016 and 2017. Dr. Villadsen stated it is clear

615 Exhibit 22570-X0562.01, PDF page 36. 616 Decision 2011-474, paragraph 68. Decision 2009-216, paragraph 255. 617 Exhibit 22570-X0193.01, PDF page 8. 618 Exhibit 22570-X0153.01, PDF page 101. 619 Exhibit 22570-X0131, PDF page 69. 620 Exhibit 22570-X0562.01, PDF pages 49, 61 and 67. 621 Decision 20622-D01-2016, paragraph 157. 622 Exhibit 22570-X0193.01, PDF page 77.



that the approved ROE both in Canada and the U.S. has been substantially higher than the ROE

awarded by the Commission in the 2016 GCOC decision. She noted that the average deemed

equity ratio is in the range of 40 to 50 per cent. Excluding Crown corporations, the approved

ROE elsewhere in Canada is approximately 9.3 per cent, and the deemed equity ratios have

averaged approximately 40 per cent.623

472. In presenting her ROE recommendations, and recognizing that the cost of equity depends

on the leverage of the company to which it is applied, Dr. Villadsen considered the difference in

leverage between the data she used to estimate the cost of equity and a benchmark equity

percentage. Using the established techniques (such as the Modigliani-Miller and Hamada

adjustments),624 Dr. Villadsen adjusted her ROE estimates for leverage and presented both the

adjusted and unadjusted recommendations. Mr. Hevert raised a similar point in his evidence.625

473. Dr. Cleary presented evidence on the relevance of market P/B ratios in assessing the cost

of equity. Dr. Villadsen submitted that consistent with her position in the 2016 GCOC, she finds

information on P/B ratios to be problematic.626 In her rebuttal evidence, Dr. Villadsen provided

further critique of Dr. Cleary’s evidence on P/B values and their relevance to the fair ROE.627

Commission findings

474. As previously discussed in Section 4, the Commission will not take any guidance from

the evidence presented about approved utility ROEs in other Canadian and U.S. jurisdictions.

The objective of the GCOC is to consider the market expectation for the affected utilities and not

what other regulators are allowing.

475. In Decision 20622-D01-2016, the Commission considered the relationship between

capital structure and ROE and techniques to account for financial risk by adjusting for leverage,

such as the Modigliani-Miller and Hamada models. The Commission concluded that “As a

consequence of the uncertainty created by the number of untested assumptions as well as the lack

of sensitivity analysis provided for some of the models, the Commission will not employ any of

these suggested models in its determination of the deemed equity ratios or the approved ROE in

this proceeding except to illustrate that a relationship exists.”628

476. The Commission has not been persuaded to depart from these earlier findings. In this

proceeding, Mr. Hevert appears to have come to a similar conclusion when he stated:

Please note that although the Modigliani-Miller and Hamada adjustments may be used to

generally measure the magnitude of the effect of incremental increases in leverage on the

Cost of Equity, it is important to recognize the results are imprecise due to the complex

and the dynamic nature of the relationship. It also is important to keep in mind that any

measure of an “optimal” capital structure must consider numerous objectives and

constraints. Nonetheless, the analytical results are consistent with the proposition that

increasing financial leverage increases the Cost of Equity.629

623 Exhibit 22570-X0193.01, PDF page 78. 624 Refer to Exhibit 22570-X0192.01, PDF pages 38-42. 625 Exhibit 22570-X0153.01, PDF pages 105-108. 626 Exhibit 22570-X0193.01, PDF page 79. 627 Exhibit 22570-X0767.01, PDF pages 72-74. 628 Decision 20622-D01-2016, paragraph 101. 629 Exhibit 22570-X0153.01, PDF pages 107-108.



477. Dr. Villadsen adjusted her overall ROE recommendation somewhat in recognition of the

Commission’s preference for results that do not take leverage into account.630

478. With respect to the relevance of P/B ratios, the Commission notes that the experts

disagreed on the merits of using these ratios in assessing the cost of equity. The Commission

further notes that no new transactions affecting Alberta utilities have been cited in evidence since

the 2013 GCOC proceeding for the Commission to consider. Consistent with the 2016 GCOC

decision,631 the Commission has not given any weight to P/B ratio evidence in this proceeding.

8.8 Conclusions on ROE

479. The Commission has been presented with a wide range of recommended ROEs as set out

in Table 6 below.

Table 6. ROE recommendations presented

Recommended by

Mr. Hevert 632 Recommended by

Dr. Villadsen633 Recommended by

Dr. Cleary634 Recommended by

Mr. Coyne635

(%)

2018 9.00 – 10.75 10.00 6.30 9.50

2019 9.00 – 10.75 10.00 6.30 9.50

2020 9.00 – 10.75 10.00 6.30 9.50

480. Mr. Hevert, on behalf of AltaLink, EPCOR and FortisAlberta, arrived at his

recommended ROE range of 9.00 to 10.75 per cent, giving primary weight to his Canadian utility

proxy group and, within that group, giving principal weight to the DCF model-based results,

with less weight given to the CAPM and risk premium based methods. Mr. Hevert submitted that

his recommendations consider higher levels of current and expected growth, increased short-term

rates, normalization of monetary policy, a continued increase in utility bond yields and increased

risk, as measured by Bloomberg’s beta coefficients.636

481. Dr. Villadsen, on behalf of the ATCO Utilities and AltaGas, recommended an approved

ROE in the range of 9.50 to 10.50 per cent, with 10.00 per cent as a reasonable point estimate.

Dr. Villadsen stated that this value was supported by her Canadian utility proxy group, her U.S.

gas LDC utility proxy group, and her U.S. water utility proxy group, before any consideration of

financial risk.637 Dr. Villadsen submitted that it was preferable to consider the ROE estimates

using multiple methods, consistent with the approach taken by other provincial regulators.

Dr. Villadsen based her recommendation on the results from her CAPM, single- and multi-stage

630 Exhibit 22570-X0193.01, PDF page 80. 631 Decision 20622-D01-2016, paragraph 305. 632 Exhibit 22570-X0153.01, PDF page 131. 633 Exhibit 22570-X0193.01, PDF page 99. 634 Exhibit 22570-X0562.01, PDF page 6. 635 Exhibit 22570-X0562.01, PDF page 117. 636 Exhibit 22570-X0153.01, PDF page 97. 637 Exhibit 22570-X0193.01, PDF page 99.



DCF models, and considering the business risk analysis of Dr. Carpenter and Mr. Buttke’s

submissions on relevant changes in global economic and Canadian capital market conditions. 638

482. Dr. Cleary, on behalf of the UCA, recommended an ROE of 6.3 per cent. In making this

recommendation, he gave equal weight to his CAPM, DCF and BYPRPM estimates. Dr. Cleary

noted that his results were reasonable compared to expected long-term market returns in the

6.00 to 9.00 per cent range, and the low-risk nature of regulated utilities.639

483. Mr. Coyne, on behalf of ENMAX, recommended an approved ROE of 9.50 per cent as a

reasonable point estimate. Mr. Coyne’s recommendation was based on the CAPM and DCF

model results for all three of his proxy groups, with greater weight placed on the results of his

North American electric proxy group and his Canadian utility proxy group.640

484. Mr. Thygesen, on behalf of the CCA, recommended an ROE of 7.75 per cent. Rather than

develop his recommendation using financial models, Mr. Thygesen compared the affected

utilities’ average actual ROE of 9.44 per cent for 2014 to 2016, to the average actual ROE of

8.90 per cent for the same period, for the companies in Mr. Hevert’s U.S. utility proxy group.

Mr. Thygesen noted the resulting difference of 54 bps and he stated that a downward adjustment

to the approved ROE for 2014 to 2016 of 8.30 per cent641 would be required to bring the ROE of

the affected utilities to the level of comparable investments.642

485. The Commission finds Mr. Thygesen’s recommended ROE, which is only based on a

comparison of average actual ROEs achieved over a three-year period between utilities in

Alberta and the U.S. is not a reasonable method to establish approved ROEs for the affected

utilities for 2018 to 2020. His comparison lacks the detailed analysis that should be performed to

identify the reasons why the actual ROEs achieved by the companies in Mr. Hevert’s U.S. utility

proxy group were different than the ROEs achieved by the affected utilities over the same period.

Accordingly, the Commission will place no weight on Mr. Thygesen’s recommendation in

determining the approved ROE.

486. Turning to the ROE estimates presented using the CAPM, the Commission found in

Section 8.2.4 that the wide range of CAPM results does not, on its own, provide much assistance

to the Commission in determining an approved ROE. Further, the relatively wide range of betas,

and interest rates still being lower relative to average historical rates, continue to be factors that

will lead the Commission to assign relatively less weight to the CAPM ROE results.

Nonetheless, the Commission has determined a point estimate of 7.90 per cent with respect to the

CAPM, which it will consider to establish an approved ROE.

487. Regarding the ECAPM, the Commission found in Section 8.2.5 that different empirical

adjustment factors may need to be employed when applied to adjusted betas or, conversely,

unadjusted betas may need to be employed in any future ECAPM that relies on the empirical

adjustment factors used by Dr. Villadsen and Mr. Hevert, and that other modifications to the

empirical ECAPM adjustment coefficients may be required, unique to regulated utilities. The

638 Exhibit 22570-X0193.01, PDF pages 11-12. 639 Exhibit 22570-X0562.01, PDF pages 74-75. 640 Exhibit 22570-X0131, PDF page 10. 641 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-019. 642 Exhibit 22570-X0551, paragraphs 120-129.



Commission remains of the view expressed in the 2016 GCOC decision that the empirical

adjustment factor in ECAPM does not resolve the issue with respect to the wide range of

estimated betas. As a result, the Commission will not assign significant weight to the ECAPM

results in this proceeding. The Commission considers it preferable to improve the CAPM results

by way of multi-factor models that specifically aim to identify factors explaining the required

return, if possible, rather than using empirical adjustment factors as is done under the ECAPM.

488. With respect to the BYPRPM, the Commission found in Section 8.3 that this approach is

a valid tool in estimating the cost of equity as it is simple to use, incorporates readily observable,

market-determined data (such as bond returns and yields), and conforms to the basic principle

that investors require a higher return for assets with greater risk. However, the BYPRP models

presented in this proceeding falter in their application of the equity risk premium adder to the

bond yield. Accordingly, the Commission did not place any weight on the results of the BYPRP

models presented by Dr. Cleary, Mr. Hevert or Mr. Coyne. The Commission did, however, take

note of Dr. Cleary’s observation that yields on Bloomberg generic long-term A-rated Canadian

utility bonds (which parties agreed track the yields on Alberta utility bonds with reasonable

accuracy) have been relatively stable since the time of the 2016 GCOC proceeding, and that this

stability in the overall yield was the result of an inverse relationship between interest rates

(which increased) and credit spreads (which narrowed) over the period leading up to this

proceeding. This led the Commission to note that changes in the interest rate and the utility bond

credit spread appear to have offset each other to some extent.

489. The Commission also found in Section 8.3 that without any meaningful analysis of the

PRPM on the record of this proceeding, and without any evidence being presented that the

PRPM has been vetted and accepted by other utility regulators as a valid approach to estimate

ROEs for regulated utilities, the Commission is not prepared to assign the PRPM any weight in

this proceeding.

490. Regarding the DCF models presented, the Commission indicated in Section 8.4 that it

preferred the multi-stage models of Dr. Villadsen and Mr. Coyne because they use forward-

looking long-term growth estimates. The Commission considers that the 8.79 per cent ROE

estimate from Mr. Coyne’s multi-stage DCF model for his North American electric proxy group,

which excludes one of the Canadian companies that had a growth rate in excess of 8.25 per cent,

and includes all of the companies from his U.S. electric proxy group, largely mitigates the issue

regarding the outcomes associated with the use of high growth rates and small proxy group sizes.

Accordingly, the Commission found that 8.79 per cent is a reasonable point estimate for the

multi-stage DCF method.

491. Regarding the evidence submitted on stock market return expectations of finance

professionals, in Section 8.5 the Commission maintained its view from previous GCOC

decisions that return expectations of finance market professionals are germane to the

determination of a fair ROE for regulated utilities, while keeping in mind the purpose and

limitations of such estimates.

492. Notwithstanding that the market return expectations of finance professionals may be of

some informational value in the determination of a fair ROE for regulated utilities, in this

proceeding the evidence received was of little assistance to the Commission for the reasons

described in Section 8.5.



493. In Section 8.7, the Commission stated that it will not take any guidance from the

evidence presented about approved utility ROEs in other Canadian and U.S. jurisdictions,

because the objective of the GCOC is to consider the market expectation for the affected utilities

in Alberta and not what other regulators are allowing. With respect to the relationship between

capital structure and ROE, and techniques to account for financial risk by adjustment for

leverage, the Commission indicated that it had not been persuaded to depart from its findings in

the 2016 GCOC decision that it would not employ any of the suggested models in its

determination of the deemed equity ratios or the approved ROE except to illustrate that a

relationship exists. In the same section, the Commission stated that it has not given any weight to

P/B ratio evidence in this proceeding.

494. In this proceeding, the Commission was presented with evidence that utility credit

spreads have narrowed since the 2016 GCOC proceeding. In Section 6, the Commission stated

that it continues to be of the view that credit spreads are an objective measure, based on

observable market data, which help to inform the Commission about utility bond investors’ risk

perceptions, and by implication, to some extent, the expectations of utility equity investors.

While evidence was put forward by Mr. Hevert that a decline in credit spreads may be of a

short-term, temporary nature, and may not be indicative of a change in risk perceptions in the

market,643 the Commission is not persuaded by this evidence, which is contradicted by

Mr. Hevert’s own claims during the 2016 GCOC proceeding that the increase in credit spreads at

that time demonstrated an increase in investors’ risk perceptions.644 The Commission agrees with

the following statement by the CCA:

The CCA does not find it credible that, co-incidental with a rise in credit spreads,

Mr. Hevert would find that credit spreads are indicative of a need to increase ROE and

when they fall, there is no correlation or that within the space of two years the

relationship would be broken. Similarly, correlation is not causation. Simply because

there is no correlation does not prove that there is no information value. It would be

difficult to argue that market has the same perception of risk when credit spreads of 206

versus 140 – it makes no sense since credit spreads are designed to compensate for risk.

Finally, while there may be no linkage in the short term, that does not prove anything for

the long term – which is what this proceeding is about. For these reasons and given the

previous findings of the Commission that credit spreads are informative as to risk, the

conclusion of AEF [AltaLink EPCOR FortisAlberta] should be accorded no weight.645

495. In Section 6, the Commission found that the global economic and Canadian capital

market conditions have improved since the 2016 GCOC proceeding, and are far removed from

the 2008-2009 financial crisis. In particular, the Commission observed that there has been global

and national economic growth, reduced market volatility, a modest increase in the 30-year GOC

bond yield and a compression in credit spreads. However, the Commission finds that the upward

pressure associated with certain of these factors is largely offset by the downward pressure

associated with others. On balance, these factors indicate the approved ROE for 2018 should be

at or near that set in the 2016 GCOC decision.

643 Exhibit 22570-X0741, PDF pages 14-18. 644 Decision 20622-D01-2016, paragraph 64. 645 Exhibit 22570-X0920, PDF page 21.



496. The Commission also found that the expectations of diminishing national GDP growth

rates, moderately higher inflation to reach the mid-point of the Bank of Canada’s target range,

increasing short-term interest rates, a flattening yield curve, but uncertain long-term interest rates

and market uncertainty with respect to international trade, result in a similar offset and together

indicate that the approved ROE for 2019 and 2020 should be the same or similar to the value set

for 2018.

497. The Commission notes that in the 2016 GCOC decision it awarded an ROE of 8.3 per

cent for 2016, and an ROE of 8.5 per cent for 2017. In its correspondence initiating this

proceeding, the Commission detailed that this proceeding would consider, amongst other things,

whether a change in the approved ROE established in the 2016 GCOC decision is warranted.

498. In the Commission’s view, if there has been some upward pressure on ROE since the

2016 GCOC proceeding, part of that pressure has already been accounted for in the 20 bps

increase in ROE awarded in 2017. No party focused on the changes since 2016 and no party

explained why this increase is either still warranted or is insufficient on a going-forward basis.

499. The 20 bps increase awarded in 2017 was premised on the Commission finding that

economic conditions were generally expected to improve in 2017, including an expected increase

in interest rates and utility bond yields. The expected increase in 30-year GOC bond yields

forecast by the witnesses in this proceeding would arguably signal an increase in approved ROE

for 2018 to 2020. However, in the Commission’s view, the concomitant expected increase in

ROE has been mitigated at least somewhat by the tightening of credit spreads. This has resulted

in utility bonds being effectively unchanged since the 2016 GCOC proceeding, contrary to what

the Commission considered would occur in Decision 20622-D01-2016. Although the

Commission cannot assume that changes in utility equity investors’ required returns will align

exactly with changes in utility bond investors’ expected return, and given that the approved ROE

has been increased by 20 bps in that same period, the Commission finds that any additional ROE

required by utility investors is largely accounted for in the 2017 adjustment approved in the 2016

GCOC decision.

500. On balance, the Commission is not persuaded by the evidence on the record that a

departure from the current approved ROE of 8.50 per cent is warranted. Consequently, the

Commission approves 8.50 per cent as the ROE for the affected utilities for 2018, 2019 and

2020.

8.9 Returning to a formula-based approach to establishing ROE

501. References were made in this proceeding to the formula-based approach to setting ROE,

previously employed by the Commission and its predecessor. For example, in response to a

question from Commission counsel with respect to the Consensus economic outlook, Mr. Coyne

stated: “… just as this Commission has done in the past when it used a formula, to look to an

outside indicator that’s readily available, it’s transparent. It takes the Commission out of the role

of having to guess what the forward path of interest rates is going to be, which is a tough

proposition.”646

646 Transcript, Volume 5, page 934



502. A standardized approach to the establishment of a single generic ROE to be applied

uniformly to all utilities and adjusted yearly using an annual adjustment formula, was approved

in Decision 2004-052. In Decision 2009-216, the Commission noted as follows:

Administrative efficiency in dealing with cost of capital evidence in rate proceedings was

clearly an impetus for the Board and parties to consider a generic ROE formula approach

and a single proceeding for setting capital structure for all utilities. The Commission

considers that the proliferation of regulated companies caused by electric and gas

deregulation, unbundling, and corporate reorganizations that influenced the Board to

adopt a generic approach remains a compelling reason to continue with that approach.647

503. While the Commission has subsequently maintained the approach of having a single

proceeding for setting a generic ROE and capital structure for all utilities, it discontinued the

annual adjustment/generic ROE formula approach in the 2009 GCOC decision. In departing

from the annual adjustment/generic ROE formula approach, the Commission accepted that

“during the current financial crisis, the traditional relationship between the risk-free rate

(measured as a yield on long Canada bonds) and the required market return on equities has not

continued.”648 The Commission also stated that it recognized “there remains a considerable amount

of uncertainty in the financial markets and the Commission is concerned that awarding a generic

ROE that does not take these uncertainties into account would be unreasonable.”649

504. The Commission understands that a formula-based approach continues to be employed

by certain other regulators. The Commission remains of the view that administrative efficiency is

an impetus for consideration of a generic ROE formula approach. The Commission also

considers that some of the issues and concerns articulated in this, and previous GCOC decisions,

in relation to the approaches to estimating ROE and the varied inputs and results, may be

remedied by adopting a formula-based approach in a future proceeding.

505. Based on the evidence regarding market conditions in this proceeding, as summarized in

Section 6, the Commission considers that returning to an annual adjustment/generic formula

approach to ROE may be reasonable. Specifically, it would appear, based on the evidence in this

proceeding, that the reasons justifying a departure from the annual adjustment formula in 2009

may no longer be a concern.

506. The Commission intends to explore the possibility of returning to a formula-based

approach to cost of capital matters. The Commission will be initiating a proceeding to explore

available options in this regard and will provide notice to that effect to all parties registered in

this proceeding in due course.

9 Capital structure matters

9.1 Overview

507. To satisfy the fair return standard, the Commission is required to determine deemed

equity ratios (also referred to as capital structure) for each of the affected utilities. In this

647 Decision 2009-216, paragraph 220. 648 Decision 2009-216, paragraph 324. 649 Decision 2009-216, paragraph 330.



decision, the Commission has established an approved ROE of 8.5 per cent for 2018 through

2020 for all of the affected utilities on a final basis.

508. For the 2018-2020 period, the Commission will maintain its previous approach of setting

a uniform approved ROE, and then adjusting for any differences in risk among each of the

affected utilities by adjusting the deemed equity ratios. The Commission will make adjustments,

if required, to recognize changes in relative risk for each affected utility from the approved

deemed equity ratios established in the 2016 GCOC decision.

509. This section of the decision determines the approved deemed percentage of rate base (net

of no-cost capital) supported by common equity. The section is organized as follows.

Section 9.2, identifies the deemed equity ratios requested by the affected utilities. The

Commission’s consideration of the factors relevant to the determination of an approved deemed

equity ratio for each affected utility begins in Section 9.3 with a review of the evidence in

relation to changes in business risk that impact all the affected utilities. In that section, the

Commission also compares business risk and deemed equity ratios between the affected utilities

and utilities in other jurisdictions. Mr. Hevert’s submissions on industry financing practices are

addressed in Section 9.4, and the submissions from FortisAlberta on capital attraction are set out

in Section 9.5.

510. Before the Commission addresses credit metrics, in Section 9.6 it examines its approach

and assesses the submissions from parties on the importance of targeting deemed equity ratios

that will permit the affected utilities to maintain A-range credit ratings. The evidence in respect

of the credit metrics required by a typical pure-play regulated utility in Canada in order to

maintain an A-range credit rating is examined in Section 9.7. In Section 9.8 and Section 9.9, the

Commission addresses whether there is a need for different deemed equity ratios for each of the

transmission, distribution and the non-taxable utilities. The Commission also evaluates the credit

metrics of the affected utilities having regard to significant financial parameters observed in

Rule 005 filings and other evidence on the record of this proceeding, including the embedded

average debt rate, depreciation as a percentage of invested capital, the income tax rate and the

mid-year construction work in progress (CWIP) as a percentage of invested capital. The

Commission addresses the submissions of ENMAX regarding its deemed equity ratio in

Section 9.10. The Commission’s approved deemed equity ratios for 2018 to 2020 for each of the

affected utilities, with the exception of AltaGas, are included in Section 9.11. The approved

deemed equity ratio for AltaGas is included in Section 10.

9.2 Deemed equity ratios requested

511. Mr. Buttke submitted that investors will look at the results of the 2018 GCOC decision to

help form their views of regulatory risk, and to discern trends.650 He stated that the reductions in

the deemed equity ratios made in the 2016 GCOC decision implied that the Commission

considered the risk of operating a utility in Alberta was decreasing. However, the market’s belief

was that the risk was increasing, based on the utility asset disposition (UAD) decision651 and the

transition to PBR.652

650 Exhibit 22570-X0179, A7. 651 Decision 2013-417: Utility Asset Disposition, Proceeding 20, Application 1566373-1, November 26, 2013. 652 Exhibit 22570-X0179, A9.



512. Dr. Villadsen653 commented that in the 2016 GCOC decision, the Commission’s focus

seemed to be on establishing a deemed equity ratio that would satisfy the bare minimum credit

quality standards necessary to obtain an A-range credit rating. Mr. Coyne stated that the

Commission appeared to have shifted away from its prior rationale for setting deemed equity

ratios on the basis of long-run business and financial risk.654 Mr. Hevert submitted that the use of

pro forma credit metrics as the basis of setting the deemed equity ratios is “concerning.”655

513. Dr. Villadsen and Mr. Coyne suggested that the credit metric analysis undertaken by the

Commission in the 2016 GCOC decision received more weight than (1) the Commission’s

finding that there was a general increase in generic business risk because of the UAD decision;

(2) the Commission’s finding that there continued to be differences in business risk as between

distribution and transmission utilities; and (3) the Commission’s acknowledgement that there is a

disadvantage for non-taxable utilities in terms of cash flow and financial flexibility.656

514. Dr. Villadsen stated that a singular focus on credit metrics is not sufficient to ensure that

a utility can (1) attract equity capital; (2) offer a return equal to that of an alternative investment

of comparable risk; and (3) provide a cushion should economic or market conditions move in a

negative direction.657

515. Dr. Villadsen indicated that her deemed equity ratio recommendations are based on her

review of commonly approved equity ratios for regulated utilities and credit metric benchmarks,

as well as a review and analysis of credit rating agencies’ commentaries on capital structures.

Dr. Villadsen recommended a 300 bps increase to the deemed equity ratios for AltaGas and the

ATCO Utilities. She stated that the resulting 40 per cent deemed equity ratio for the ATCO

Utilities is consistent with other regulated utilities in Canada.658

516. AltaLink659 and EPCOR660 suggested that while a consideration of credit metrics is

important, it is only one factor to be considered in determining a fair return, because credit

metrics do not properly account for relevant business risk, financial risks and uncertainties.

They submitted that quantitative and qualitative business risk factors must be taken into account,

similar to how credit rating agencies take both into account when determining credit ratings.661

517. FortisAlberta stated that the Commission’s exercise of judgment in determining capital

structure should be expanded to address other important factors relating to the role that deemed

equity ratios play in overall capital attraction, including the importance of ensuring that equity

investors remain willing to support the utility’s operations.662

518. Mr. Hevert explained that his recommended deemed equity ratio focuses on industry

financing practices and ongoing business risks, and considers the Commission’s practice of

653 Exhibit 22570-X0193.01, A75. 654 Exhibit 22570-X0131, PDF page 82. 655 Exhibit 22570-X0153.01, PDF page 111. 656 Exhibit 22570-X0193.01, A78. Exhibit 22570-0131, PDF page 97. 657 Exhibit 22570-X0193.01, A75 and A79. 658 Exhibit 22570-X0193.01, A5. 659 Exhibit 22570-X0141, paragraphs 27 and 32. 660 Exhibit 22570-X0195, paragraph 59. 661 Exhibit 22570-X0141, paragraph 44. Exhibit 22570-X0195, paragraphs 59 and 61. 662 Exhibit 22570-X0228, paragraphs 14-15.



referring to certain credit metrics.663 After considering industry financing practices, the historical

variability in the parameters underlying the pro forma credit metric calculations used by the

Commission in the 2016 GCOC decision, the breadth of data considered by credit rating

agencies in arriving at credit ratings, and the relationship between financial leverage and the cost

of equity, Mr. Hevert recommended a deemed equity ratio of 40 per cent for AltaLink, EPCOR

and FortisAlberta for 2018, 2019 and 2020.664

519. Mr. Coyne considered the differences in business risk as between the affected utilities

and the utilities in his U.S. electric proxy group, in combination with financial risks, in

recommending a deemed equity ratio of 40 per cent for the taxable Alberta electric transmission

and distribution utilities. He recommended that the Commission restore the 200 bps adder for the

non-taxable utilities in Alberta to compensate for their reduced cash flows and weaker credit

metrics. Consequently, his recommended deemed equity ratio for ENMAX is 42 per cent.665

520. Calgary argued that circumstances have not changed enough for the Commission to

change either the ROE or deemed equity ratios approved in the 2016 GCOC decision. However,

Calgary submitted that if the Commission determines that changes should be made, the deemed

equity ratio for ATCO Gas should be reduced to 35 per cent, as Mr. Johnson concluded that its

business risk is at the low end for natural gas and electricity distribution companies in Canada.666

521. Mr. Madsen considered that the Commission’s past practice of using credit metrics to

assess the deemed equity ratios remains appropriate, and ensures that the approved deemed

equity ratios support the ability of the affected utilities to continue to operate in a safe, reliable

and economic manner.667 Mr. Madsen indicated that the primary focus of his recommendations

on deemed equity ratios was a consideration of credit metrics and matters relevant to those credit

metrics.668 Mr. Madsen performed an assessment of each utility to arrive at his recommended

deemed equity ratio for that utility.669 His assessment included a review of the deemed equity

ratio he calculated for each utility to achieve an A-range credit rating, and a consideration of any

utility specific risks. Mr. Madsen’s recommended deemed equity ratios are set out in Table 7

below.

522. Dr. Cleary commented that the Alberta utilities possess low risk, as demonstrated by their

low earnings volatility, their ability to generate high operating profit margins, and their

opportunity for growth in operating earnings. Based on these considerations, combined with his

positive economic and capital market outlook, Dr. Cleary recommended no change in the

deemed equity ratios. Dr. Cleary instead emphasized the impetus for a reduction in the approved

ROE. He submitted that his recommendations are supported by the credit metric analysis

provided by Mr. Bell.670

523. Mr. Bell recommended a deemed equity ratio of 37 per cent, which he submitted is well

within the credit metric guidelines established by S&P and DBRS Limited (DBRS) to maintain

663 Exhibit 22570-X0153.01, PDF page 7. 664 Exhibit 22570-X0153.01, PDF page 11. 665 Exhibit 22570-X0131, PDF pages 9, 87-88, 101. 666 Exhibit 22570-X0903, PDF pages 4-5. Exhibit 22570-X0611.02, A4. 667 Exhibit 22570-X0557, paragraph 144. 668 Exhibit 22570-X0557, paragraph 118. 669 Exhibit 22570-X0557, paragraph 265. 670 Exhibit 22570-X0562.01, PDF page 6.



an A-range credit rating. He did not object to the continuation of a 400 bps increase in deemed

equity ratio for AltaGas.671

524. The currently approved deemed equity ratios, and the recommended figures for 2018,

2019 and 2020, are set out in the following table.

Table 7. Currently approved deemed equity ratios and the deemed equity ratios recommended for 2018, 2019 and 2020

Last approved

672

Recommended by AltaGas and

the ATCO

Utilities673

Dr. Villadsen

Recommended by AltaLink/ EPCOR/

FortisAlberta674

Mr. Hevert

Recommended

by ENMAX675

Mr. Coyne

Recommended

by Calgary676

Mr. Johnson

Recommended

by the CCA677

Mr. Madsen

Recommended by

the UCA678

Dr. Cleary

(%)

Electricity and natural gas transmission

AltaLink 37 40 35 37

ATCO Electric Transmission

37 40

35 37

ATCO Pipelines 37 40 36 37

ENMAX Transmission 36 42 36 36

EPCOR Transmission 37 40 36 37

Lethbridge 37

Red Deer 37

TransAlta 37

Electricity and natural gas distribution

AltaGas 41 44 41 41

ATCO Electric Distribution

37 40

36 37

ATCO Gas 37 40 35 35 37

ENMAX Distribution 36 42 36 36

EPCOR Distribution 37 40 36 37

FortisAlberta 37 40 35 37

671 Exhibit 22570-X0559, A18. 672 Decision 20622-D01-2016, Table 26, paragraph 622. For ATCO Electric Transmission, the deemed equity ratio

was approved in Decision 22121-D01-2016: ATCO Electric Ltd. Transmission Operations, Application for

Finalization of Return on Equity and Deemed Equity Ratio for 2016-2017, Proceeding 22121, December 16,

2016. For ENMAX Transmission and ENMAX Distribution, the deemed equity ratio was approved in Decision

22211-D01-2017: ENMAX Power Corporation, Application for Finalization of Deemed Equity Ratio for 2016-

2017, Proceeding 22211, July 27, 2017. 673 Exhibit 22570-X0193.01, Figure 33. 674 Exhibit 22570-X0153.01, PDF page 123. 675 Exhibit 22570-X0131, PDF page 10. 676 Exhibit 22570-X0611.02, PDF page 2. 677 Exhibit 22570-X0557, PDF page 74. 678 Exhibit 22570-X0562.01, PDF page 96. Transcript, Volume 10, pages 2098-2099.



9.3 Generic business risk analysis

525. In this section of the decision, the Commission considers the evidence on business risk

factors impacting all the affected utilities, or a particular segment of the affected utilities, that

may require the Commission to adjust the deemed equity ratios approved in the 2016 GCOC

decision.

526. As previously mentioned, Mr. Hevert, Mr. Coyne, Mr. Johnson and Dr. Cleary each

indicated that business risk was either one of the factors, or the primary factor, underlying their

recommended deemed equity ratios. In addition, based primarily on the business risk assessment

undertaken by Dr. Carpenter, Dr. Villadsen argued for using the deemed equity ratios of U.S.

utilities as comparators.679

527. Dr. Carpenter defined business risk as “the underlying risks inherent in a particular

company’s operations.” He added that while business risk is “a somewhat subjective concept,

and there is more than one way of structuring an analysis of business risk, an approach that is

commonly taken is to consider five elements of business risk: supply risk, demand (or market)

risk, competitive risk, operating risk and regulatory risk.”680 Mr. Hevert agreed that these five

elements of business risk all have a direct bearing on earnings levels and volatility.681

528. Dr. Carpenter assessed the business risk of AltaGas and the ATCO Utilities relative to

their business risk in the past, and relative to the business risks of utilities in other jurisdictions.

He particularly focused on utilities owned by the companies that Dr. Villadsen used as proxy

groups in her evidence. Dr. Carpenter’s analysis also focused on the natural gas and electricity

distribution functions, which he noted the Commission had used as a benchmark in prior

proceedings.682

529. Mr. Coyne undertook a proxy group risk analysis in order to help determine his

recommended equity ratios. Noting the limited number of companies in his Canadian utility

proxy group, Mr. Coyne looked to a U.S. sample of low-risk electric utilities. Mr. Coyne

indicated that he examined the business and financial risks of his U.S. electric proxy group,

relative to those of a typical Alberta electric transmission and electric distribution utility.683

530. Mr. Johnson assessed the business risk of ATCO Gas relative to other natural gas

distributors in Canada and the U.S.

531. Mr. Hevert identified uncertainties associated with regulation that are faced by AltaLink,

EPCOR and FortisAlberta.

532. Dr. Cleary primarily used quantitative analysis to assess the business risks of the utilities

in Alberta on an overall basis, as well as in comparison to U.S. utilities.

533. The Commission will first examine the overall assessment of business risk of the utilities

in Alberta offered by Dr. Cleary. Next, the Commission will address changes in business risks

679 Exhibit 22570-X0193.01, A17. 680 Exhibit 22570-X0186, A10. 681 Exhibit 22570-X0153.01, PDF page 21. 682 Exhibit 22570-X0186, A4. 683 Exhibit 22570-X0131, PDF page 86.



since the 2016 GCOC decision that were identified by Dr. Carpenter, Mr. Coyne, Mr. Hevert and

the affected utilities. Subsequent to that, the Commission will address the business risk

comparisons between the affected utilities and other jurisdictions submitted by Dr. Carpenter,

Mr. Coyne, Mr. Johnson and Dr. Cleary.

9.3.1 Overall assessment of business risk

534. Dr. Cleary agreed with the favourable assessment of business risk for the affected utilities

included in credit rating reports issued by DBRS and S&P.684 Dr. Cleary stated that regulated

Alberta operating utilities possess low business risk and enjoy solid regulatory support.685

Dr. Cleary undertook some empirical analysis that purported to support his conclusion that the

affected utilities operate in a low-risk environment that enables them to earn above their

approved ROEs with very little volatility in income.686

535. Part of Dr. Cleary’s empirical analysis examined the ability of the affected utilities to

earn their approved ROE on a consistent basis from 2005 to 2016, which he described as a

bottomline measure of the total risks faced by the utilities.687 The yearly figures illustrated that

the affected utilities earned average and median ROEs above the approved ROE in all years

except 2005, when the average ROE was 0.18 per cent below the approved ROE. Dr. Cleary

submitted this can be considered the strongest indication that the affected utilities possess low

overall risk.688

Commission findings

536. The Commission accepts that the favourable financial performance and low volatility of

earnings illustrated by Dr. Cleary is support for the conclusion that the affected utilities have

generally low business risk.

9.3.2 Changes in business risk since the 2016 GCOC decision

537. The affected utilities and their witnesses focussed on issues related to regulatory risk. The

main issues identified were (1) the 2018-2022 PBR term; (2) the Commission’s UAD decision

and the related issue of asset utilization; (3) the increase in customer contributions; (4) regulatory

lag; and (5) clean energy initiatives. These issues will be addressed in the following sections.

9.3.2.1 2018-2022 PBR term

538. Dr. Carpenter, Mr. Coyne and EPCOR submitted that changes associated with the

2018-2022 PBR term will increase risk, primarily with respect to the distribution utilities’ ability

to recover operating and capital costs.689 ENMAX noted the Commission’s adoption of the K-bar

methodology, and submitted that the Commission’s willingness to reopen aspects of the PBR

framework at the last minute, and in isolation, is a troubling development that materially

increases the risks and uncertainty that its distribution utility faces.690

684 Exhibit 22570-X0562.01, PDF pages 74-75. 685 Exhibit 22570-X0562.01, PDF page 75. 686 Exhibit 22570-X0562.01, PDF page 77. 687 Exhibit 22570-0562.01, PDF page 79. 688 Exhibit 22570-0562.01, PDF pages 79-81. 689 Exhibit 22570-X0186, A41. Exhibit 22570-X0131, PDF page 78. Exhibit 22570-X0733, A22. 690 Exhibit 22570-X0773, paragraphs 17-21.



539. With respect to the increased uncertainty of operating cost recovery, Dr. Carpenter

submitted that the Commission’s approach to rebasing for the 2018-2022 PBR term does not

align revenues with costs at the start of the term; as a result, business risk is increased.691

Dr. Carpenter stated it is unusual that rebasing is not cost-of-service based,692 and he commented

that this is a fundamentally different approach for rebasing.693 Mr. Coyne agreed that the

2018-2022 PBR term plan is a significant departure from the previous PBR plan, and provides

additional risk on several fronts.694 Dr. Carpenter submitted that the Commission’s decision to

reject all of the expense anomalies proposed by the distribution utilities will create a shortfall for

the utilities and increases business risk.695 He proposed that rejection of the anomalies requested

by AltaGas and the ATCO Utilities amount to a reduction of 50 bps in annual ROE.696

540. Dr. Carpenter submitted that the Commission’s evolving approach to the calculation of

K-bar, including its decision to apply a new K-bar methodology for the 2018-2022 PBR plan, is

a source of increased risk.697 He submitted this new methodology creates a disconnect between

the need for and the availability of supplemental capital funding, which creates capital recovery

risk.698 EPCOR submitted that the annual updating of the K-bar will reduce the certainty and

predictability of the capital funding that will be provided.699 Dr. Carpenter estimated the impact

of applying this annual K-bar update for ATCO Electric Distribution and ATCO Gas to be

equivalent to an annual reduction in authorized ROE of over 100 bps.700 EPCOR indicated that

annual updates to base K-bar alone would reduce its expected ROE by nearly 200 bps.701

541. EPCOR suggested there is substantial uncertainty as to whether its distribution function

will have access to the supplemental funding it requires to address the AESO’s proposed 2018

tariff application, which will require more transmission connection projects costs to be funded by

customer contributions.702 Mr. Coyne referred to the provincial government’s target of 30 per

cent renewable generation by 2030, which the government hopes to accomplish in part by the

widespread deployment of distributed generation. Mr. Coyne stated that costs required for the

distribution utilities to develop infrastructure capable of integrating increased volumes of

distributed generation are currently not provided for under PBR.703

542. The UCA submitted that the use of a K-bar mechanism under the 2018-2022 PBR term

improves the incentive properties of the PBR plan. It stated that any changes made to the K-bar

mechanism in the 2018-2022 PBR term rebasing decision do not impact the underlying PBR plan

and it suggested that the annual updating of the inputs into the K-bar mechanism could result in

increased K-bar revenues.704 The UCA stated that the Commission addressed concerns about lack

691 Exhibit 22570-X0751, A64. 692 Exhibit 22570-X0751, A61. 693 Exhibit 22570-X0186, A41. 694 Exhibit 22570-X0775, PDF page 58. 695 Exhibit 22570-X0751, A59-A60. 696 Exhibit 22570-X0751, A63. 697 Exhibit 22570-X0751, A51. 698 Exhibit 22570-X0751, A52. 699 Exhibit 22570-X0733, A11. 700 Exhibit 22570-X0751, A66, A69. 701 Exhibit 22570-X0733, A11. 702 Exhibit 22570-X0195, paragraphs 23, 25, 27-28. 703 Exhibit 22570-X0131, PDF page 76. 704 Exhibit 22570-X0767.01, paragraph 298.



of sufficient funding under K-bar, as part of Decision 22394-D01-2018. It noted that the

Commission was not persuaded by the arguments of the utilities in that proceeding about having

a lack of sufficient funding.705

543. Mr. Bell disagreed with the concerns about the 2018-2022 PBR term raised by the

utilities and their experts. He submitted that the financial performance of the distribution utilities

under the first PBR term improved, when compared to the utilities that remained under cost of

service. Mr. Bell indicated that the actual ROEs for the distribution utilities under the first term

of the PBR plan improved, as compared to their actual ROEs prior to PBR. He submitted this

improvement indicates that risk declined in the first PBR term. Mr. Bell noted that while the

approved ROEs have declined from the time prior to the first PBR term, the actual ROEs

achieved over the first PBR term have increased.706

544. Mr. Johnson stated that being under a PBR regime does not increase the regulatory risk of

ATCO Gas. He referred to the actual ROEs earned by ATCO Gas in 2013 (11.86 per cent), 2014

(10.95 per cent), 2015 (11.10 per cent) and 2016 (12.93 per cent), and noted that, in each year,

the actual ROEs were in excess of the approved ROEs, which were 8.30 per cent.707

545. Dr. Carpenter submitted that Mr. Bell’s examination of historically achieved ROEs is

unlikely to be meaningful because it compares actual ROEs averaged over a different number of

years. Based on his own calculations using data averaged over time periods of four years,

Dr. Carpenter reported that the difference between the actual ROEs in the most recent four-year

PBR time period and the prior four-year cost-of-service period are nearly the same as the

difference in the actual ROEs between the 2005 to 2008 and 2009 to 2012 cost-of-service

periods.708 EPCOR commented that Mr. Bell’s use of four data points lacks statistical rigour.709

546. EPCOR commented that increased returns under PBR are not surprising because of the

incentives that exist under PBR, and they are not indicative of decreased risk. EPCOR submitted

it will have more difficulty identifying and implementing efficiency improvements during the

2018 to 2022 PBR term and because of this, it will face greater uncertainty and risk under the

2018 to 2022 PBR term than it did under the first PBR term.710

547. Mr. Madsen stated that the distribution utilities will have a reasonable opportunity and

incentives to recover their prudently incurred costs over the 2018 to 2022 PBR term, which is

consistent with the Commission’s findings in Decision 20414-D01-2016 (Errata).711

548. The UCA submitted that the intent of PBR was not to increase risk, but rather to provide

appropriate incentives for regulated utilities to improve efficiencies and share any resulting cost

savings with customers. It contended that this intent will be significantly undermined if the

utilities are able to successfully argue that the presence of such incentives increases their risk and

705 Exhibit 22570-X0913, paragraph 164. 706 Exhibit 22570-X0559, A14. 707 Exhibit 22570-X0611.02, A7. 708 Exhibit 22570-X0751, A46. 709 Exhibit 22570-X0733, A7. 710 Exhibit 22570-X0733, A6. 711 Decision 20414-D01-2016 (Errata): Errata to Decision 20414-D01-2016, 2018-2022 Performance-Based

Regulation Plans for Alberta Electric and Gas Distribution Utilities, Proceeding 20414, February 6, 2017.

Exhibit 22570-X0557, paragraph 127.



therefore requires additional compensation through the GCOC. The UCA stressed that in the

2013 GCOC decision, the Commission was not persuaded that the transition to PBR had resulted

in a change in the risk profile that warranted any adjustments to the approved ROE and deemed

equity ratios.712

Commission findings

549. As part of the 2013 GCOC decision, the Commission considered whether PBR increased

the business risk of the distribution utilities. In that decision, the Commission denied a request

for an increase of 75 bps in the deemed equity ratio because of the implementation of PBR.713

The Commission noted that the utilities had the opportunity to apply for Y, Z and K factor

adjustments and, since the implementation of PBR in 2013, all but one of the distribution utilities

achieved actual ROEs in 2013 that were in excess of the interim approved ROE for that year.

The Commission also noted that the risks asserted by the distribution utilities had not manifested

themselves through credit rating downgrades.714

550. In this proceeding, the evidence likewise fails to support that any changes associated with

the 2018 to 2022 PBR term will increase risk to the distribution utilities, including risk

respecting the ability to recover operating and capital costs. As a preliminary observation, the

Commission notes that while the 2018 to 2022 PBR term includes adoption of new rebasing and

capital funding mechanisms, the underlying structure and intent of PBR is largely unchanged.

The Commission has received no persuasive evidence of any negative market response to the

PBR framework since its implementation in 2013. The Commission agrees with the UCA’s

submission that the intent of implementing PBR was not to increase risk for the distribution

utilities, but to provide incentives for the utilities to improve efficiencies, and to benefit

financially from these improvements. The evidence supports that intent has generally been

realized. While EPCOR submitted it will have more difficulty identifying and implementing

efficiency improvements during the 2018 to 2022 PBR term, this does not rule out the possibility

that efficiencies will continue to be achieved. In addition, the 2018 to 2022 PBR plan retains the

opportunity to apply for Y, Z and K factor adjustments, and includes off-ramp and reopener

provisions to safeguard the financial integrity of the affected utilities.

551. As for the more specific risks asserted by Mr. Coyne, Dr. Carpenter and EPCOR with

respect to the distribution utilities’ ability to recover operating and capital costs due to changes

associated with the 2018-2022 PBR term, the Commission has, in Decision 20414-D01-2016

(Errata), stated it “is satisfied that the distribution utilities will have a reasonable opportunity to

earn their allowed rates of return over the next generation PBR plans.”715 The Commission added

to this finding in Decision 22394-D01-2018 as follows:

425. The Commission is satisfied that the distribution utilities will have a reasonable

opportunity to earn their allowed rates of return over the period covered by the 2018-

2022 PBR plans. The Commission has reached this conclusion having had regard to the

evidence filed in Proceeding 20414, the additional evidence filed in this compliance

proceeding and the elements of the PBR plans approved in Decision 20414-D01-2016

712 Exhibit 22570-X0913, paragraphs 156-157. 713 Decision 2191-D01-2015, paragraph 380. 714 Decision 2191-D01-2015, paragraphs 377-378. 715 Decision 20414-D01-2016 (Errata), paragraph 288.



(Errata) as refined in this decision, and having applied its experience, expertise and

judgement in carrying out its mandate to set just and reasonable rates.716

552. No persuasive evidence has been offered to support a conclusion contrary to those quoted

above from Decision 20414-D01-2016 (Errata) and Decision 22394-D01-2018.

553. Based on the foregoing, the Commission finds that there is no increase in business risk as

a result of the 2018 to 2022 PBR plan.

9.3.2.2 The Commission’s UAD decision and the related issue of asset utilization

554. In the 2016 GCOC decision, the Commission addressed the issue of incremental business

risk to utility investors stemming from developments with respect to the UAD decision that

occurred between the 2013 GCOC proceeding and the 2016 GCOC proceeding.717 In the current

proceeding, Mr. Buttke indicated that the market took some comfort from the Commission’s

acknowledgement in the 2016 GCOC decision that the UAD decision directionally increased

investor risk.718

555. AltaLink719 and EPCOR720 submitted that no new developments have occurred since the

2016 GCOC decision that would reduce their risk associated with the UAD decision.

Dr. Carpenter,721 EPCOR,722 FortisAlberta723 and AltaLink724 pointed out that since the 2016

GCOC decision, new developments have occurred with respect to the UAD decision that

increase the business risk of the affected utilities.

556. Dr. Carpenter noted several changes since the 2016 GCOC decision. He noted that the

Government of Alberta initiated a consultation with stakeholders regarding possible legislation

to address outstanding concerns in relation to UAD-related decisions. He submitted this

underscores the significance of investor uncertainty associated with the Commission’s UAD

policy.725 EPCOR and FortisAlberta agreed that this consultation has increased the level of

uncertainty regarding their ability to recover capital.726

557. Dr. Carpenter also noted that the Commission expressed its intent to initiate a process to

consider the issue of transmission asset utilization, which was raised by the CCA as part of

deferral account proceedings for AltaLink and ATCO Electric Transmission.727 AltaLink stated

that any suggestion that a denial of prudently incurred capital costs could be based on asset

utilization is a significant new uncertainty that will be closely monitored by capital market

716 Decision 22394-D01-2018, paragraph 425. 717 Decision 20622-D01-2016, Section 7.4.1. 718 Exhibit 22570-X0179, A13. 719 Exhibit 22570-X0141, paragraph 14. 720 Exhibit 22570-X0195, paragraph 21. 721 Exhibit 22570-X0186, A61. Exhibit 22570-X0751, A66, A69, A76. 722 Exhibit 22570-X0195, paragraph 21. 723 Exhibit 22570-X0228, paragraph 28. 724 Exhibit 22570-X0141, paragraph 13. 725 Exhibit 22570-X0186, A61. 726 Exhibit 22570-X0195, paragraph 21. Exhibit 22570-X0228, paragraph 28. 727 Exhibit 22570-X0186, A5.



participants. AltaLink indicated that credit rating agencies have already taken notice of this

issue.728

558. Dr. Carpenter indicated his understanding that the Commission’s asset utilization

proceeding will raise the prospect that capital cost recovery might be denied for certain prudently

constructed new electric transmission capital assets. He submitted this represents a new source of

capital recovery risk for the electric transmission utilities.729 Dr. Carpenter stated he is not aware

of any other regulatory jurisdictions that are considering whether cost recovery for prudently

incurred capital assets should be denied.

559. Dr. Carpenter submitted that the future uncertainty about potential disallowances is what

causes an increase in business risk. He noted that in its decision on the 2018 to 2022 PBR term

rebasing,730 which was issued since the 2016 GCOC decision, the Commission denied EPCOR’s

request to recover the undepreciated capital costs of its conventional meters.731 Dr. Carpenter

submitted that while the costs at stake with respect to EPCOR’s conventional meters were

relatively small, the decision was important because it reaffirmed the Commission’s application

of the UAD principles, and emphasized the Commission’s belief that it lacks any discretion in

their application.732

560. EPCOR stated that the asset utilization proceeding has increased the level of uncertainty

regarding its ability to recover invested capital.733 AltaLink suggested that where this issue will

end up is unclear, but it is clear that it increases the uncertainty that must be assessed by any

informed investor in an Alberta utility.734

561. The CCA noted that AltaLink was acquired by Berkshire Hathaway Energy after the

issuance of the UAD decision. It submitted this is clear evidence that sophisticated shareholders

are aware of, and accept, the UAD risks in Alberta.735

562. Mr. Madsen commented that the Commission has already compensated the affected

utilities for the UAD asset risk.736 He also contended that any risks associated with UAD are fully

or partially offset by the upside benefits obtained when assets are sold.737

563. Mr. Bell suggested that the return associated with UAD risk for a utility should be the

gains realized when a utility sells a capital asset, and he stated that the affected utilities have

benefitted from removing property from rate base, in the amount of $17.3 million at least, to

date.738

728 Exhibit 22570-X0141, paragraphs 9, 11. 729 Exhibit 22570-X0186, A57, A60. 730 Decision 22394-D01-2018. 731 Exhibit 22570-X0751, A74-A75. 732 Exhibit 22570-X0751, A66, A69, A76. 733 Exhibit 22570-X0195, paragraph 21. 734 Exhibit 22570-X0141, paragraph 13. 735 Exhibit 22570-X0920, paragraph 256. 736 Exhibit 22570-X0557, paragraph 177. 737 Exhibit 22570-X0557, paragraph 182. 738 Exhibit 22570-X0559, A9-A10.



564. Mr. Johnson stated that, unlike other Canadian regulators, the Commission has

enunciated UAD principles, which confirm that the affected utilities have an equal opportunity to

enhance their return by selling capital assets that are no longer required to provide utility service.

Mr. Johnson submitted that this additional return offsets the potential for losses. He contended

that ATCO Gas has been successful in this regard.739

565. Dr. Carpenter contended that the risk of disallowances under UAD is a forward-looking

risk and because of this, any past gains earned by the affected utilities cannot reduce the

forward-looking risk.740 EPCOR noted that Mr. Bell’s submissions were filed before the release

of the Commission’s decision on the 2018 to 2022 PBR term rebasing, in which the Commission

denied EPCOR’s request to recover $9 million related to its conventional meters.741

566. AltaLink submitted it is critical to be proactive in managing risk. It contended that it

takes many years to recover from a credit rating downgrade when one occurs. AltaLink

commented that the Commission has made it clear that it is more beneficial to act to mitigate risk

than to wait and take action after a downgrade has occurred.742

567. The CCA pointed out that, as described in Decision 20407-D01-2016,743 EPCOR made a

decision to proceed with its advanced metering infrastructure project regardless of the

Commission’s determinations on the treatment of the remaining net book value of the

conventional meters.744 It stated that the Commission also made a number of findings of fact in

Decision 20407-D01-2016, which made it clear that there would be stranded asset costs at the

time of the PBR rebasing.745

568. The UCA commented that the Commission first addressed EPCOR’s meter replacement

issue in Decision 3100-D01-2015,746 which was issued in January 2015. It noted that in Decision

3100-D01-2015, the Commission determined that the book value of the undepreciated meters

was to the account of EPCOR’s shareholder.747 The UCA summarized that in Proceeding 22394,

EPCOR argued that the law had changed to give the Commission greater flexibility and

discretion to depart from the strict application of the principles applied in the Stores Block

decision. The UCA indicated that in Decision 22394-D01-2018, the Commission disagreed that

such discretion exists and found no basis upon which to alter its previous findings.748

569. Mr. Madsen stated that asset utilization is a risk that is related to the UAD decision, and

he noted that the Commission’s asset utilization proceeding had not been initiated as of January

2018.749 He submitted that no real cost, or risk of a cost, has materialized to date with regard to

739 Exhibit 22570-X0611.02, A8. 740 Exhibit 22570-X0751, A82. 741 Exhibit 22570-X0733, A30. 742 Exhibit 22570-X0738, paragraphs 54, 57-58. 743 Decision 20407-D01-2016: EPCOR Distribution & Transmission Inc., 2014 Capital Tracker True-Up and 2016-

2017 PBR Capital Tracker Forecast, Proceeding 20407, February 7, 2016. 744 Decision 20407-D01-2016, paragraph 647. 745 Decision 20407-D01-2016, paragraphs 616-618. 746 Decision 3100-D01-2015: EPCOR Distribution & Transmission Inc., 2013 PBR Capital True-Up and 2014-

2015 PBR Capital Tracker Forecast, Proceedings 3216 and 3100, Applications 1610565-1 and 1610362-1,

January 25, 2015. 747 Decision 3100-D01-2015, paragraph 691. 748 Exhibit 22570-X0913, paragraph 51. 749 Exhibit 22570-X0557, paragraph 177.



asset utilization.750 The CCA submitted the asset utilization issue has existed at least since the

issuance of the UAD decision, and therefore it is not a new issue.751

570. On May 8, 2018, AltaGas and the ATCO Utilities, in the cover letter accompanying their

reply argument, noted that Bill 13: An Act to Secure Alberta’s Electricity Future, had been tabled

in the Alberta legislature on April 19, 2018. AltaGas and the ATCO Utilities stated that Bill 13

touched on a number of matters that received attention at the hearing, and noted that legislative

action was specifically identified in Dr. Carpenter’s evidence regarding UAD. AltaGas and the

ATCO Utilities stated that they were reserving their rights to address the impact of any

legislative changes on the cost of capital over the 2018 to 2020 period and requested that the

Commission confirm that the record would be reopened to address the impact of any legislative

changes over the GCOC test period.

Commission findings

571. The three factors cited by the affected utilities in support of their submission that their

business risk has increased since the 2016 GCOC decision as a result of the UAD decision or the

related issue of asset utilization are (1) the Government of Alberta’s stakeholder consultation;

(2) the Commission’s intent to initiate a process to consider transmission asset utilization; and

(3) the Commission’s decision to deny EPCOR’s request to recover the undepreciated capital

costs of its conventional meters. The Commission addresses each of these considerations below.

572. The Alberta legislature passed Bill 13 on June 11, 2018. It does not include any

provisions relating to the UAD decision, stranded asset cost recovery or asset utilization. Nor has

any party requested that the record of this proceeding be reopened to address the impact of these

legislative changes, as discussed in the above-mentioned correspondence filed on behalf of

AltaGas and the ATCO Utilities on May 8, 2018. Accordingly, the Commission is not persuaded

that the outcome of the Government of Alberta’s stakeholder consultation on the UAD decision

has resulted in any change in business risk for the affected utilities for the 2018 to 2020 period.

573. With respect to the transmission asset utilization issue, the Commission advised parties

on June 20, 2017, that it would be “issuing a bulletin shortly to initiate a process to consider the

issue.”752 As of the close of the record of this proceeding, no bulletin has been issued and no

proceeding has been initiated. The timing and outcome of any transmission asset utilization

proceeding that may be subsequently held is unknown at this time and purely speculative.

574. Additionally, issues around transmission asset utilization are connected to the UAD

decision, “and how the corporate and property law principles applied by the courts in the Alberta

legislative context as referenced in the UAD decision may relate.”753 The Commission considers

that the markets and investors have had ample opportunity to become familiar with these

principles since the UAD decision was released in November 2013. The UAD decision and

subsequent decisions implementing its principles are not new, and the Commission expects that

investors already factor this into their decision making. The CCA noted that Berkshire Hathaway

Energy acquired AltaLink subsequent to the issue of the UAD decision.

750 Exhibit 22570-X0557, paragraph 179. 751 Exhibit 22570-X0920, paragraph 316. 752 Exhibit 22393-X0150, paragraph 8. 753 Exhibit 22393-X0150, paragraph 7.



575. For the foregoing reasons, the Commission finds that speculation regarding the potential

outcome of any future asset utilization proceeding is not justification for an overall increase in

business risk.

576. As to the suggestion that business risk has increased as a result of the Commission’s

denial of EPCOR distribution’s request to recover $9 million related to its conventional meters,

the Commission agrees with the CCA and the UCA that this issue was initially addressed in

2015, when the Commission issued Decision 3100-D01-2015. In that decision, the Commission

determined that the book value of the undepreciated conventional meters was to the account of

EPCOR’s shareholder.754 In Decision 22394-D01-2018, which was issued on February 5, 2018,

the Commission found no basis upon which to alter its previous findings of fact with respect to

this matter.755 The Commission finds that the confirmation of previous findings made with

respect to UAD in decisions from 2015 and early 2016 are not reflective of an increase in

business risk for the affected utilities since the 2016 GCOC proceeding.

577. In conclusion, the Commission is not satisfied that there has been an increase in business

risk for the affected utilities since the 2016 GCOC proceeding with regard to the UAD decision

or the related issue of asset utilization.

9.3.2.3 Increase in customer contributions

578. AltaLink stated that it has been, and continues to be, exposed to risks and liabilities

associated with owning and operating capital assets for which customer contributions have been

received, but on which it earns no return. It noted that these customer contributions continue to

increase, and there is the potential for these to increase even more because of the AESO’s

proposal in its 2018 tariff application that will classify more transmission connection project

costs to be funded by customer contributions.756 EPCOR indicated that this potential increase in

customer contributions also creates uncertainty for its transmission function.757

579. EPCOR stated that its transmission function is responsible for the operating and

maintenance (O&M) costs for any capital assets funded by customer contributions, and it faces

forecasting risk with respect to these O&M costs. EPCOR contended that forecasting risk is

typically compensated by the return component of the revenue requirement, but in the case of

customer contributions, it receives no return and thus no compensation for the forecasting risk.758

580. Mr. Madsen stated that AltaLink’s forecast balance of the gross and net customer

contributions as of December 31, 2017 and December 31, 2018, as a percentage of the gross and

net property, plant and equipment (PP&E) balances, do not exceed historical actual levels.759

581. Mr. Bell commented that the vast majority of the customer contributions received by the

electricity transmission utilities are from the electricity distribution utilities. He suggested that if

the transmission utilities are correct about their level of risk increasing because of increased

754 Decision 3100-D01-2015, paragraph 691. 755 Decision 22394-D01-2018, paragraph 395. 756 Exhibit 22570-X0141, paragraphs 23-24. 757 Exhibit 22570-X0195, paragraph 32. 758 Exhibit 22570-X0195, paragraph 31. 759 Exhibit 22570-X0557, paragraphs 223-227.



customer contributions, then there should be a corresponding decrease in the risk for the

distribution utilities.760

582. EPCOR commented that any customer contributions paid from its distribution utility to

its transmission utility increases uncertainty and risk to both entities. It explained that the

increased uncertainty for the distribution utility arises because the recovery of its carrying costs

related to the customer contributions is not guaranteed under the PBR framework.761

Commission findings

583. AltaLink’s forecast customer contribution amounts at the end of 2018 comprise less than

10 per cent of its total PP&E. The forecast percentages for 2018 (9.0 per cent of the gross PP&E

and 9.4 per cent of the net PP&E) are within the range of the actual percentages for the years

2013 to 2016, which range from 8.3 per cent to 9.1 per cent of the gross PP&E, and 8.7 per cent

to 10.4 per cent of the net PP&E.762 This evidence does not support AltaLink’s submissions

regarding increased business risk since the 2016 GCOC proceeding.

584. While both AltaLink and EPCOR referenced the potential for electricity transmission

utilities to receive increased customer contributions because of proposals included in the AESO’s

2018 tariff application, no decision on the AESO’s 2018 tariff application has been issued. The

Commission cannot know whether this proposal will be accepted or not, and it will not speculate

on the outcome. An AESO proposal not yet addressed is not justification for an increase in

business risk.

585. EPCOR submitted that there would be increased business risk for its distribution utility if

the AESO’s proposal is approved, because of the uncertainty associated with the capital funding

mechanism under the 2018 to 2022 PBR plan. The Commission has previously found that the

2018 to 2022 PBR plan does not increase the business risk of the distribution utilities relative to

the risk at the time of the 2016 GCOC proceeding.

586. No evidence was presented that would enable the Commission to assess whether the risks

in operating assets that were funded in whole or in part by customer contributions are any

different than the risks in operating assets for which no customer contributions have been

received. No evidence was presented on how customer contributions may help reduce stranded

asset risk.

587. Customer contributions are treated as no-cost capital, as are funds collected for FIT. The

pre-collection of future returns through CWIP-in-rate base, and the pre-collection of funds

through higher salvage rates and excess depreciation rates, can also be considered a form of no-

cost capital on regulatory balance sheets.

588. The Commission allowed CWIP-in-rate base and the collection of FIT for AltaLink and

ATCO Electric Transmission to assist with cash flow and credit metric support during the large

transmission build, but has allowed these utilities to refund those no-cost capital accumulations.

ATCO Electric Transmission had requested an increase in net salvage in their last general tariff

application (GTA) but was denied. AltaLink had requested an increase in net salvage in their last

760 Exhibit 22570-X0559, A19-A23. 761 Exhibit 22570-X0733, A26. 762 Exhibit 22570-X0464, AML-CCA-2017NOV21-005.



litigated GTA before their negotiated settlement GTA and that was partially approved. In this

GCOC proceeding, AltaLink indicated that they are looking at potentially applying for a

reduction in net salvage on the same terms as EPCOR.763

589. The Commission observes that no-cost capital is an inherent aspect of regulated utilities’

balance sheets and requests for increases and decreases to these no-cost capital balances have

occurred over the years for many reasons. Cash flow injections have assisted utilities at critical

times in their operations and have supported credit metrics. Yet at the same time utilities request

increases to their approved ROE because of the increase in business risk of managing these no-

cost capital assets.

590. With CWIP-in-rate base removed, and pre-collected FIT amounts as well as pre-collected

excess depreciation amounts refunded by AltaLink, in the Commission’s view it can be argued

that utilities have reduced their business risk through reductions in no-cost capital.

591. For all of the above reasons, the Commission finds that there is no increase in business

risk from customer contributions for the affected utilities since the 2016 GCOC proceeding.

9.3.2.4 Regulatory lag

592. Dr. Carpenter stated that approximately $3 billion of capital costs are subject to deferral

account proceedings for the electric transmission utilities, and the lag in finalizing these

proceedings creates uncertainty.764 Mr. Buttke commented that regulatory lag generates increased

uncertainty with respect to revenues and ROE. He added that this increased uncertainty will

cause investors to increase their required rate of return, all else equal, or it will cause them to

shift their capital to jurisdictions with less uncertainty.765

593. AltaLink stated that it remains subject to regulatory lag. It noted that the 2018 GCOC

decision will result in prospective ROEs for all of 2019 and all of 2020, but only for a portion of

2018. It anticipates a decision on its 2014 direct assigned capital deferral account (DACDA) in

2018 at the earliest, and it advised that this decision will not result in final approval if the asset

utilization issue is not resolved by that time. AltaLink stated that it has approximately $4 billion

of completed capital projects pending prudency reviews through DACDA proceedings.766

594. AltaLink commented that regulatory lag is harmful because it creates market uncertainty

and increases the risk of adverse credit-rating action. It added that regulatory lag increases the

uncertainty and volatility in cash flows, which increases the market perception of risk.767

595. Mr. Madsen contended that if the Commission disallows any capital expenditures made

by a utility, it would be because the expenditure was deemed to be imprudent. He submitted that

any such disallowances should not be considered when the Commission determines a fair return

for the utilities.768

763 Transcript, Volume 6, page 1088. 764 Exhibit 22570-X0131, A5. 765 Exhibit 22570-X0179, A10. 766 Exhibit 22570-X0141, paragraphs 15, 17-18. 767 Exhibit 22570-X0141, paragraphs 19 and 21. 768 Exhibit 22570-X0557, paragraphs 204-207.



596. AltaLink submitted that it is not seeking compensation for prudency risk. It explained

that regulatory lag prevents the timely implementation of ongoing Commission findings and

recommendations into its project execution, in order to address any prudency concerns the

Commission has identified.769 AltaLink also argued that increased regulatory lag increases the

risk of adverse credit-rating action.

Commission findings

597. The Commission agrees with Dr. Carpenter and Mr. Buttke that regulatory lag creates

uncertainty. However, this lag has existed for many years and is not new. The 2016 GCOC

decision resulted in an approved ROE and deemed equity ratios that were fully prospective for

one year. The ROE and deemed equity ratios approved by the Commission in this decision will

result in fully prospective ROE and deemed equity ratios for two full years, which the

Commission considers will help reduce regulatory lag related to the cost of capital element of

rates for these years.

598. Mr. Buttke and AltaLink argued that regulatory lag increases uncertainty with respect to

revenues and cash flows. While AltaLink and Dr. Carpenter noted the large amount of capital

additions that are subject to deferral account proceedings for the electric transmission utilities,

this is not reflective of the cash balances in the deferral accounts that these utilities have

requested as part of those deferral account proceedings. No information was provided on the

magnitude of these balances, without which the Commission cannot properly assess the cash

flow and revenue uncertainty associated with these deferral account proceedings.

599. The Commission acknowledges that the capital additions the electric transmission

utilities have requested be added to rate base as part of their deferral account proceedings are

substantial. These capital additions will be assessed for prudence, as is the Commission’s normal

practice. The Commission is not persuaded, however, that the magnitude of capital additions

currently in a deferral account presents a different level of risk than at the time of the 2016

GCOC proceeding. The Commission has addressed the argument for any potential increased risk

associated with these deferral account proceedings because of the asset utilization issue in

Section 9.3.2.2.

600. Based on the foregoing, the Commission considers that regulatory lag has, in general,

stayed the same or improved relative to the period leading up to the 2016 GCOC decision.

Accordingly, the Commission does not find that business risk has increased since the 2016

GCOC proceeding as a result of regulatory lag.

9.3.2.5 Clean energy initiatives

601. Dr. Carpenter indicated that policies to encourage the connection of distributed

generation could reduce utilization of some electric transmission assets in the future.770

Mr. Coyne commented that recent clean energy initiatives encourage utility customers to pursue

distributed generation, and this will reduce customer demand.771 He added that the expansion of

769 Exhibit 22570-X0891, paragraph 48. 770 Exhibit 22570-X0186, A60. 771 Exhibit 22570-X0131, PDF page 73.



distributed generation could significantly impact the long-term business risk profile of the

distribution utilities.772

602. EPCOR suggested that there is an increasing likelihood that Alberta will see growing

levels of distributed generation in the near term and further into the future. It stated that the

prospect of these increased levels of distributed generation creates uncertainty for utility

investors because of the possibility of stranded assets, and the potential for increased costs that

are not contemplated within the PBR plan.773

Commission findings

603. The issue of the impact of green energy initiatives was raised by Mr. Hevert in the 2016

GCOC proceeding. In the Commission’s view, this is not an entirely new development. Given

the minimal information provided in this proceeding with respect to the actual and forecast levels

of distributed generation and associated impacts on the distribution systems, the Commission is

not in a position to adequately assess the effect that clean energy initiatives will have on the

long-term business risk profile of the affected utilities. The Commission was in the same position

in the 2016 GCOC proceeding. The Commission is not persuaded that the clean energy

initiatives that have been instituted since the 2016 GCOC proceeding have increased the business

risk of the affected utilities.

604. The Commission has addressed the issue of stranded assets in connection with the UAD

decision in Section 9.3.2.2. The Commission has addressed the issue of capital cost recovery and

availability under the 2018 to 2022 PBR term in Section 9.3.2.1. In both of those sections, the

Commission found that there was no increase in business risk for these two elements since the


9.3.3 Business risk comparisons between the affected utilities and other jurisdictions

605. Dr. Carpenter stated that the Commission did not comment on the similarity of business

risk of utilities in the U.S. and Canada in the 2016 GCOC decision.774 In this section, the

Commission will address the comparisons made as part of this proceeding by Dr. Carpenter,

Mr. Coyne, Mr. Johnson and Dr. Cleary, and consider the relative regulatory risk in the U.S. and

Alberta.

606. Dr. Carpenter assessed the business risk of AltaGas and the ATCO Utilities relative to

their business risk in the past, and relative to the business risks of utilities in other jurisdictions.

He focused particularly on utilities owned by the companies that Dr. Villadsen used as proxy

groups in her evidence. Dr. Carpenter’s analysis also focused on the natural gas and electricity

distribution functions, which he noted the Commission had used as a benchmark in prior

proceedings.775 Based primarily on the business risk assessment undertaken by Dr. Carpenter,

Dr. Villadsen argued for using the deemed equity ratios of U.S. utilities as comparators.776

772 Exhibit 22570-X0131, PDF page 74. 773 Exhibit 22570-X0195, paragraph 22. 774 Exhibit 22570-X0186, A24. 775 Exhibit 22570-X0186, A4. 776 Exhibit 22570-X0193.01, A17.



607. Mr. Coyne undertook a proxy group risk analysis in order to help determine his

recommended equity ratios. Noting the limited number of companies in his Canadian utility

proxy group, Mr. Coyne looked to a U.S. sample of low-risk electric utilities. Mr. Coyne

indicated that he examined the business and financial risks of his U.S. electric proxy group,

relative to those of a typical Alberta electric transmission and electric distribution utility.777

608. Dr. Carpenter stated that the regulatory risks facing distribution utilities in Alberta and

the U.S. are similar, and both are relatively low risk. He indicated that both the U.S. regulatory

regime and the Alberta regulatory regime are supportive in relation to long-term capital cost

recovery, with the exception of the Commission’s UAD policy.778

609. Dr. Carpenter noted the Commission’s concerns in previous GCOC proceedings that in

comparing the regulatory frameworks between the U.S. and Canada, there are differences due to

the use of deferral accounts and reduced regulatory lag in Canada. Dr. Carpenter submitted that

if he were to compare a jurisdiction that makes significant use of deferral accounts with a

jurisdiction that does not, he would expect the jurisdiction that uses deferral accounts to have

slightly lower business and regulatory risk, but the difference would not be large.779

610. Dr. Carpenter indicated that none of the utilities in Dr. Villadsen’s U.S. gas LDC utility

proxy group are exposed to commodity price risk, most of them have some form of revenue

decoupling, and most have a capital tracker mechanism.780 Dr. Carpenter expected regulatory lag

to be a relatively minor contributor to business risk differentials, unless regulatory lag gives rise

to a risk that invested capital will not be recovered. He noted the regulatory lag in Alberta

associated with the electric transmission capital asset deferral account proceedings, and he

indicated that the asset utilization proceeding will be reviewing electric transmission capital asset

additions from 2014 onward.781

611. Noting changes associated with the 2018 to 2022 PBR term, as discussed in Section

9.3.2.1 above, Dr. Carpenter indicated he is not aware of any distribution utilities in the U.S. that

are exposed to these risks. He stated that these regulatory risk factors significantly differentiate

the utilities in Alberta from the utilities in Dr. Villadsen’s U.S. gas LDC utility proxy group.782

612. Dr. Carpenter submitted that the business risk of AltaGas and the ATCO Utilities are

similar to those of the utilities in Dr. Villadsen’s U.S. gas LDC utility proxy group, with the

exception of UAD risk and PBR risk.783 He submitted that since the 2016 GCOC proceeding,

business risk in Alberta has increased because of the Commission’s decisions on the 2018 to

2022 PBR term and the asset utilization issue.784

613. Dr. Carpenter indicated he would not expect there to be large differences in business risk

between the companies in Dr. Villadsen’s U.S. gas LDC utility proxy group and the companies

in her U.S. water utility proxy group. Dr. Carpenter considered the companies in Dr. Villadsen’s

777 Exhibit 22570-X0131, PDF page 86. 778 Exhibit 22570-X0186, A16. 779 Exhibit 22570-X0186, A35. 780 Exhibit 22570-X0186, A36. 781 Exhibit 22570-X0186, A37. 782 Exhibit 22570-X0186, A48. 783 Exhibit 22570-X0186, A31. 784 Exhibit 22570-X0186, A62.



U.S. pipeline proxy group to constitute an upper bound of the risk that a natural gas distribution

utility might face, primarily because the pipeline companies have higher business risk due to

greater exposure to competition risk and more regulatory risk.785

614. Based on Dr. Carpenter’s identification of the UAD risk and the PBR risk for AltaGas

and the ATCO Utilities, which he submitted the U.S. natural gas distribution utilities do not face,

Dr. Carpenter judged the business risk of AltaGas and the ATCO Utilities to be greater than the

risk of Dr. Villadsen’s U.S. gas LDC utility proxy group, but not as great as Dr. Villadsen’s U.S.

pipeline proxy group.786 Dr. Villadsen agreed.787

615. Mr. Coyne summarized his comparison of business risk between the utilities in Alberta

and the companies in his U.S. electric proxy group as follows:

In sum, I find risk profiles of the U.S. proxy group and the Alberta utilities to be different

but comparable. The U.S. proxy group has somewhat more risk due to the vertical

integration of its utilities. But, Alberta utilities are directly exposed to changes in

throughput due to declining load or loss of customers, whereby nearly half the U.S.

utilities are protected from such risks through decoupling mechanisms. Both jurisdictions

have established regulatory processes geared towards providing reasonably timely cost

recovery and mitigating regulatory lag through the use of forecast test years and capital

trackers, though Alberta’s recent use of historical OM&A [operating, maintenance and

administration] data and capital for rebasing its PBR plan is a significant departure from

established precedents. Consequently, Alberta utilities have greater risk under a multi-

year PBR plan where costs and revenues are deliberately decoupled. The reliance on a

PBR framework in Alberta places earnings at greater risk and adds risk relative to the

U.S. utilities that are predominantly regulated on a cost of service basis. Alberta utilities

also have greater risk due to the low level of awarded returns in Alberta and the

uncertainty around cost recovery stemming from the UAD Decision. These risks do not

exist elsewhere in the proxy group. Overall, I consider the U.S. proxy group to have

lower risk than the Alberta utilities, despite the added risk to the U.S. proxy group for its

inclusion of vertically integrated electric utilities, and will consider these risk differences

in combination with financial risks in recommending an equity ratio for Alberta’s electric

transmission and distribution utilities.788

616. Mr. Coyne submitted there is no substance to the belief that U.S. electric utilities are

measurably riskier than the affected utilities. He noted an upgrade made by Moody’s to most

U.S. utilities in January 2014 to reflect its revised view that U.S. regulators have generally

provided regulated utilities a reasonable opportunity to recover costs and returns.789

617. Mr. Coyne submitted details of the regulatory environments under which the companies

in his U.S. electric proxy group operate. The 33 operating companies in his U.S. electric proxy

group are primarily regulated electric transmission and distribution utilities. Of the 33 operating

companies, six of them operate in two states and one operates in three states. The information

provided by Mr. Coyne included the jurisdictions the companies operate in, the jurisdiction’s

regulatory risk assessments, the regulatory framework under which the companies operate, the

785 Exhibit 22570-X0186, A50-A51. 786 Exhibit 22570-X0186, A50-A51. 787 Exhibit 22570-X0193.01, A17. 788 Exhibit 22570-X0131, PDF pages 87-88. 789 Exhibit 22570-X0775, PDF page 55.



test year basis, whether there is revenue decoupling and the parent companies credit rating.

While Mr. Coyne indicated that regulatory lag for these companies is mitigated by the use of

deferral accounts,790 no information was provided on the nature of these deferral accounts,

including whether the operating companies have deferral accounts associated with capital project

costs.791

618. Mr. Coyne summarized that (1) the majority of the companies in his U.S. electric proxy

group operate under regulatory frameworks that are based on costs of service, in exclusive

territories; (2) more than half of the companies operate under a forecast or partial forecast test

year; (3) the parent companies have an average credit rating of A-; and (4) the companies operate

in regulatory jurisdictions that are ranked slightly above the average for the constructive nature

of the regulatory environment. He noted that many of the companies are vertically integrated,

and he stated there is somewhat more risk because of this.

619. The information provided by Dr. Carpenter as part of his regulatory risk comparison was

not as detailed as the information provided by Mr. Coyne. Dr. Carpenter focused on (1) whether

there was a revenue decoupling mechanism in the states where the companies in Dr. Villadsen’s

U.S. gas LDC utility proxy group operate; (2) whether there was a capital tracker mechanism in

the states these companies operate in; and (3) providing information about the rate case dates.

Dr. Carpenter did not indicate whether all the companies were holding companies, operating

companies or some combination of the two.792

620. Mr. Johnson’s assessment of relative business risk was restricted to ATCO Gas. He

submitted that, unlike most other natural gas distribution companies in Canada and the U.S.,

ATCO Gas has one of the lowest supply risks. Mr. Johnson noted that, unlike Union Gas and

Enbridge Gas in Ontario, ATCO Gas has a weather deferral account that protects it from reduced

consumption. Mr. Johnson commented that ATCO Gas has minimal market risk, and has similar

regulatory risk to the other utilities in Alberta. Mr. Johnson identified that the operating risk for

ATCO Gas may have increased because the urban main pipelines it proposes to acquire from

ATCO Pipelines have not been tested for integrity.793

621. Dr. Cleary’s analysis of business risk centered on numerical factors. Dr. Cleary used a

CV of the earnings before interest and income taxes (EBIT)/sales ratio to quantify the level of

business risk of the affected utilities and a number of the U.S. utilities used by Dr. Villadsen,

Mr. Hevert and Mr. Coyne in their evidence. Based on his analysis, Dr. Cleary stated that the

affected utilities have less volatility in operating profit margins, which demonstrates lower

business risk than the U.S. utilities.794

622. Dr. Cleary also compared the affected utilities to the U.S. utilities on the basis of the CV

of their earned ROEs from 2005 to 2016. Dr. Cleary concluded that the U.S. utilities displayed

790 Exhibit 22570-X0131, PDF page 87. 791 Exhibit 22570-X0132, worksheet JMC-9 Regulatory Risk. 792 Exhibit 22570-X0186, A36-A38. 793 Exhibit 22570-X0611.02, A6. 794 Exhibit 22570-X0562.01, PDF pages 82-86.



much greater volatility in ROEs than the affected utilities, which again suggests that the U.S.

utilities possess greater risk than the affected utilities.795

623. Dr. Carpenter disagreed with the use of historical accounting-based ROEs to assess

business risk in the context of setting the ROE and the deemed equity ratios. He submitted that

any comparison involving historical ROEs does not constitute evidence of business risk on a

go-forward basis.796

624. Mr. Coyne submitted that the EBIT/sales ratio represents a company’s profit margin, but

not its earnings. He stated that operating profits are measured by EBIT.797 He stated that

Dr. Cleary’s use of the EBIT/sales ratio to compare U.S. and Canadian utilities should be

dismissed because it is not related to business risk, but rather revenue mix. He contended that

revenue mix is not a factor in discussing the variability of earnings.798 Dr. Carpenter noted that

the sales figures for the U.S. utilities that Dr. Cleary used in his analysis of the CV of the

EBIT/sales ratios include commodity revenues, whereas this is not the case for the affected

utilities.799

625. Dr. Carpenter noted Dr. Cleary’s concern about using an analysis of the CV of the EBIT

of the affected utilities, in the context of the high rate base growth of the affected utilities over

the last 10 years. Dr. Carpenter calculated an alternative measure that subtracts out the impact of

growth from the CV (EBIT), and indicated that the earnings volatility from 2005 to 2016 for the

affected utilities is comparable to the U.S. utilities that Dr. Cleary analyzed through his CV of

the EBIT/sales ratio.800

626. Dr. Villadsen pointed out some inconsistencies in Dr. Cleary’s CV of ROE comparison.

She noted that Dr. Cleary’s analysis excludes all of the companies in her U.S. pipeline proxy

group and her U.S. water utility proxy group, among other U.S. companies, as well as a group of

publicly traded Canadian utility holding companies. Dr. Villadsen stated that Dr. Cleary’s

analysis used data from 2005 to 2017 for the affected utilities, but used data from 2007 to 2016

for the U.S. utilities.801

627. Dr. Villadsen stated that the companies in her U.S. gas LDC utility proxy group and her

U.S. water utility proxy group have a much lower CV of ROE than the U.S. companies used by

Dr. Cleary in his analysis.802 Dr. Villadsen added that these lower CVs of ROE are similar to

those of the affected utilities. She noted, however, that the affected utilities have earned lower

returns on average than either her U.S. gas LDC utility proxy group or her U.S. water utility

proxy group.803

628. Instead of quantifying volatility based on the CV (EBIT/sales) calculation that Dr. Cleary

used, Mr. Hevert stated that the CV of net operating income (NOI) was a more appropriate

795 Exhibit 22570-X0562.01, PDF pages 89-91. 796 Exhibit 22570-X0751, A90. 797 Exhibit 22570-X0775, PDF page 49. 798 Exhibit 22570-X0775, PDF page 51. 799 Exhibit 22570-X0751, A18. 800 Exhibit 22570-X0751, A20. 801 Exhibit 22570-X0767.01, A16. 802 Exhibit 22570-X0767.01, A19. 803 Exhibit 22570-X0767.01, A19.



measure of business risk because income taxes are an operating expense for utility companies.

He commented that the affected utilities have the highest CV of NOI. Mr. Hevert stated that the

CV (NOI), together with the CV of earned ROE, is another measure regarding relative riskiness.

Mr. Hevert submitted that the average of the CV (NOI) and the CV (ROE) shows that all proxy

groups considered by the parties in this proceeding are relevant in deriving an ROE for the

affected utilities.804 Dr. Cleary questioned the informative value of averaging these two ratios.805

629. Mr. Hevert also described S&P’s use of the volatility of profitability, when S&P weighs

profitability in its assessment of financial risk. Mr. Hevert noted that when S&P’s approach is

applied to Dr. Cleary’s data, it demonstrates that the Canadian utilities’ EBIT and earnings

before interest, income taxes, depreciation and amortization (EBITDA) margins are not less

volatile than the U.S. utilities. Based on this, Mr. Hevert stated he does not agree with

Dr. Cleary’s claim that the U.S. utilities display greater operating income variability.806

630. Mr. Buttke submitted that equity analysts focus on differences in outcomes across

regulatory jurisdictions. He noted a recent equity research article from Canadian Imperial Bank

of Commerce (CIBC) in which CIBC articulated its view that differences in the U.S. and

Canadian regulatory environments may make U.S. acquisitions attractive for Canadian utilities.807

AltaLink referred to a report from February 2016 in which DBRS scored the regulatory regime

for electric transmission utilities in Alberta as below average with respect to deemed equity

percentages, political interference and stranded cost recovery.808

Commission findings

631. In the 2009 GCOC decision, the Commission agreed that the business risks, other than

regulatory risks, of the utility business are similar as between utilities in Alberta and the U.S.809

Based on the evidence presented during this proceeding, the Commission remains of this view.

632. With respect to regulatory risk, the Commission considered in the 2009 GCOC decision

that while the differences in regulatory practice between the U.S. and Canada may have

narrowed, on the whole, “Canadian utilities enjoy a more supportive regulatory environment and

have less regulatory risk than their American counterparts.”810

633. In this proceeding, both Dr. Carpenter and Mr. Coyne took the position that companies in

certain U.S. proxy groups have lower regulatory risk than the utilities in Alberta. While

Dr. Carpenter stated that the regulatory risks facing distribution utilities in Alberta and the U.S.

are similar, he judged that because of the UAD and PBR risks the Alberta utilities face, their

regulatory risk is higher than the companies in Dr. Villadsen’s U.S. gas LDC utility proxy group.

Mr. Coyne submitted that the companies in his U.S. electric proxy group have lower regulatory

risk than the affected utilities.

804 Exhibit 22570-X0741.01, PDF pages 58-59. 805 Transcript, Volume 10, page 2057. 806 Exhibit 22570-X0741.01, PDF pages 59-60. 807 Exhibit 22570-X0179, A15. 808 Exhibit 22570-X0141, paragraph 43. 809 Decision 2009-216, paragraph 144. 810 Decision 2009-216, paragraph 168.



634. Dr. Carpenter focused on the Commission’s prior identification of differences between

the U.S. regulatory regime and Canada with regard to the use of deferral accounts and regulatory

lag. Regarding deferral accounts, Dr. Carpenter indicated that most of the companies in

Dr. Villadsen’s U.S. gas LDC utility proxy group have some form of revenue decoupling and

most have a capital tracker mechanism. The Commission observes that in three of the states in

which these companies operate, the operations are limited to less than 10 per cent of the

company’s total rate base.811 Of the remaining nine states, three of them have no revenue

decoupling, five have partial decoupling through the use of weather normalization, and one

accounts for differences between authorized and actual revenues, except for the effects of

weather.812 The average annual revenue for these six companies is $1.7 billion.813 Compared to

ATCO Gas, which had revenue of approximately $1 billion in 2016814 and has a weather deferral

account, the Commission considers that the companies in Dr. Villadsen’s U.S. gas LDC utility

proxy group face much more revenue risk.

635. Dr. Carpenter mentioned the use of capital tracker mechanisms in the jurisdictions where

the companies in Dr. Villadsen’s U.S. gas LDC utility proxy group operate. The Commission

notes that three of the nine states have no such mechanism.815 For the six states that utilize capital

tracker mechanisms, no information was submitted with respect to the capital funding that is

provided through these mechanisms, compared to what has been provided for the distribution

utilities in Alberta. Consequently, the Commission is not able to make an informed assessment of

the value of the capital tracker mechanisms as between Alberta and these six states.

636. The Commission agrees with the submission of Dr. Carpenter that regulatory lag for the

distribution utilities in Alberta is similar to that of the six companies used by Dr. Villadsen in her

U.S. gas LDC utility proxy group. The period between rate cases of five years for the Alberta

distribution utilities, as noted by Dr. Carpenter, is equivalent to the average years between rate

cases for the companies in Dr. Villadsen’s U.S. gas LDC utility proxy group.816 The Commission

is aware that the distribution utilities in Alberta will have their K-bar funding mechanism

updated annually. No evidence was provided regarding the frequency of the capital tracker

mechanism approval for the companies in Dr. Villadsen’s U.S. gas LDC utility proxy group.

637. Based on its review above of the information provided by Dr. Carpenter about the

regulatory jurisdictions under which the companies in Dr. Villadsen’s U.S. gas LDC utility proxy

group operate, the Commission finds Dr. Carpenter’s submission that the regulatory risks facing

distribution utilities in Alberta and the U.S. to be similar is unsupported.

638. The Commission considers that the wide variation in practice with respect to revenue

decoupling and capital trackers does not establish any type of consistent baseline that would

support Dr. Carpenter’s submission that the regulatory risks facing distribution utilities in

Alberta and the U.S. are similar. In addition, Dr. Carpenter’s analysis did not address some of the

other differences identified in the 2009 GCOC decision, including (1) the increased importance

811 Exhibit 22570-X0186, Table 1. 812 Exhibit 22570-X0186, Table 2. 813 Exhibit 22570-X0193.01, Figure 10. 814 Exhibit 22570-X0163.01, PDF page 96. 815 Exhibit 22570-X0186, Table 3. 816 Exhibit 22570-X0186, A39 and Table 4.



in the U.S. of “the reliance of market forces as a substitute for hands on regulation,”817 which led

to unexpected consequences and an unexpected exposure to business risk; (2) the use of forward

test years in Canada compared to their use in the U.S.; and (3) a review of depreciation studies

when stranded asset risk changes.818

639. The Commission also reviewed the information provided by Mr. Coyne about the

regulatory jurisdictions under which the companies in his U.S. electric proxy group operate.

Based on the review of this information, the Commission finds Mr. Coyne’s submission that the

regulatory risks for the companies in his U.S. electric proxy group are lower than the utilities in

Alberta, to be unsupported.

640. The 33 operating companies in Mr. Coyne’s U.S. electric proxy group operate in 27

different states. The regulatory systems in place for these 27 states are not consistent. Nine have

regulatory regime rankings of above average, 15 have rankings of average, and three are ranked

as being below average. The regulatory frameworks in the 27 states are (1) original cost, which is

used in 18 states; (2) known and measurable adjustments, which is in place for five states;

(3) average rate base, for two of the states; (4) fair value, in one state; and (5) alternative rate

plans, used in one state. The test year methodologies in place are (1) fully forecasted, for 13

states; (2) historical, for 10 states; (3) partially forecasted, for two states; and (4) two states that

use both fully forecasted and historical.819 The Commission considers that the wide variation in

regulatory rankings, regulatory frameworks and test year methodologies among the 27 states

does not establish any type of consistent baseline that would support Mr. Coyne’s submission

that the regulatory risks are lower for the companies in his U.S. electric proxy group than they

are for the utilities in Alberta.

641. Six of the 33 companies in Mr. Coyne’s U.S. electric proxy group operate in two states,

while another operates in three states. Of the six companies that operate in two states, five of

them are faced with regulatory regimes that are not entirely consistent. The company that

operates in three states faces three different test year methodologies. The Commission considers

that the regulatory risk faced by the companies that operate in multiple states, under multiple

regulatory regimes, is greater than that faced by the utilities in Alberta.

642. Similar to that of Dr. Carpenter’s analysis, Mr. Coyne’s analysis did not address the

increased importance in the U.S. of “the reliance of market forces as a substitute for hands on

regulation,”820 which led to unexpected consequences and an unexpected exposure to business

risk.

643. Dr. Carpenter and Mr. Coyne commented that the utilities in Alberta face a unique risk

with respect to the UAD decision. The Commission recognized this in the 2016 GCOC decision

when it determined that regulatory risk for investors in Alberta utilities had increased by some

incremental but unquantifiable amount as a result of the Stores Block-UAD line of decisions.821

Dr. Carpenter and Mr. Coyne also indicated that the rebasing and capital funding mechanism

817 Decision 2009-216, paragraph 152. 818 Decision 2009-216, Section 3.2.2.1, Section 3.2.2.3. 819 Exhibit 22570-X0132, worksheet JMC-9 Regulatory Risk. 820 Decision 2009-216, paragraph 152. 821 Decision 20622-D01-2016, paragraph 521.



risks faced by the distribution utilities in Alberta are not faced by U.S. utilities. The Commission

has addressed this in Section 9.3.2.1.

644. With respect to Dr. Cleary’s quantification of the differences in the business risks

between the utilities in Alberta and the U.S., by calculating the CV of the EBIT/sales ratios, and

the CV of the earned ROEs, the Commission agrees with the submissions of Dr. Carpenter,

Mr. Coyne and Mr. Hevert that the EBIT/sales ratio is not valid for determining the volatility of

operating income. The Commission also agrees with Dr. Villadsen that the CV (ROE)

comparison that Dr. Cleary undertook contained inconsistencies.

645. From a quantitative perspective, the Commission takes note that (1) Dr. Carpenter’s CV

(EBIT) analysis indicated comparability between the utilities in Alberta and the U.S. companies

analyzed by Dr. Cleary; (2) Dr. Villadsen’s CV (ROE) analysis indicated similarity between the

utilities in Alberta and the companies in her U.S. gas LDC utility proxy group and her U.S. water

utility proxy group; (3) Mr. Hevert’s analysis indicated that the utilities in Alberta have the

highest CV (NOI); (4) Mr. Hevert’s use of the average of the CV (NOI) and CV (ROE) indicated

comparability between the utilities in Alberta and the U.S. utilities; and (5) Mr. Hevert’s

volatility of profitability analysis indicated that the utilities in Alberta are no less volatile than

the U.S. utilities.

646. The Commission considers that there is no single accepted mathematical way to quantify

business risk, as demonstrated by the number of different quantitative analyses undertaken by the

parties in this proceeding,

647. Based on the determinations above, the Commission finds there is no basis to support the

proposal that regulatory risk for U.S. utilities is lower than it is for the utilities in Alberta. The

Commission is also satisfied that, for the reasons expressed above, the Commission’s conclusion

in the 2009 GCOC decision still holds; that is, “while the differences in regulatory practice

between the U.S. and Canada may be narrower,”822 on the whole, “Canadian utilities enjoy a

more supportive regulatory environment and have less regulatory risk than their American

counterparts.”823

9.3.4 Comparability of deemed equity ratios

648. As previously mentioned, one of Dr. Villadsen’s considerations in the determination of

her recommended deemed equity ratio was a review of commonly approved equity ratios for

regulated utilities, including those of U.S. utilities.

649. Dr. Villadsen indicated that the 37 per cent deemed equity ratio approved in the 2016

GCOC decision is several hundred bps lower than the average for other regulated utilities in

Canada, and much lower than the ratios for distribution and transmission utilities in the U.S.

Dr. Villadsen noted that the approved ROEs for other regulated utilities in Canada and the U.S.

are higher than those in Alberta. She suggested that where the utilities in Alberta have lower

ROEs and capital structures than their counterparts, the comparability standard is only satisfied if

822 Decision 2009-216, paragraph 168. 823 Decision 2009-216, paragraph 168.



the utilities in Alberta have significantly lower business risk. Dr. Villadsen submitted that this is

not the case, based on Dr. Carpenter’s evidence on business risk.824

650. Mr. Coyne indicated that the deemed equity ratios of 36 and 37 per cent awarded for

2016 and 2017, respectively, when combined with the approved ROE of 8.5 per cent, results in

weighted equity returns of 3.06 and 3.15 per cent. He stated that these weighted returns are the

lowest for comparably regulated electric utilities in all jurisdictions across Canada, with a few

exceptions.825

651. Mr. Buttke noted DBRS’s view that the deemed equity ratio of 37 per cent awarded in the

2016 GCOC decision was in the below-average category.826

652. Dr. Villadsen submitted that a benchmark deemed equity ratio of at least 40 per cent,

before any company specific adjustments, would be necessary to place the approved equity

returns in the range of comparability relative to other regulated distribution and transmission

utilities.827 Dr. Villadsen commented that while the deemed equity ratios recommended by the

interveners range from 35 to 37 per cent, the average deemed equity ratios most recently

approved in Canada were 36.59 per cent for electricity distributors and 39.86 per cent for natural

gas distributors.828

653. Using the average approved ROEs and deemed equity ratios for (1) Canadian electricity

distributors; (2) Canadian natural gas distributors; and (3) U.S. natural gas distributors,

Dr. Villadsen calculated the resulting return on a $1 million rate base. She did the same

calculation using the ROE and deemed equity ratios recommended by the UCA, Calgary and the

CCA in this proceeding. The differences in the resulting returns between the use of approved

figures and the use of the figures recommended by the three interveners ranged from 16 to 37 per

cent when compared to the Canadian average, and from 41 to 53 per cent when compared to the

U.S. natural gas distributors. Dr. Villadsen stated she did not see any evidence that suggests

AltaGas and the ATCO Utilities should receive an ROE that is 16 to 37 per cent lower than that

granted for other Canadian utilities.829

654. Mr. Coyne noted that the deemed equity ratios for the companies in his U.S. electric

proxy group are significantly higher than those in Canada. He suggested this difference is

explained, in part, by the different processes used by Canadian and U.S. regulators for setting

equity ratios. Mr. Coyne consequently did not recommend any adjustment to account for the

different equity ratios between Canadian utilities and the companies in his U.S. electric proxy

group.830

655. The UCA noted the Commission’s previously expressed preference for an approach to

estimating the cost of capital that relies on sound principles of finance, as opposed to simply

looking to the awards of other regulators developed on the basis of different records and under

different circumstances. It noted and agreed with the observation of the chair of this proceeding

824 Exhibit 22570-X0193.01, A82. 825 Exhibit 22570-X0131, PDF page 82. 826 Exhibit 22570-X0179, A13. 827 Exhibit 22570-X0193.01, A83. 828 Exhibit 22570-X0767.01, A5. 829 Exhibit 22570-X0767.01, A98. 830 Exhibit 22570-X0131, PDF page 87.



that relying on the returns approved by other regulators necessarily imports circularity into the

process. The UCA also agreed with the chair that the relevant consideration is the market

expectation of the cost of capital, and not what other regulators are allowing.831

Commission findings

656. With respect to the comparability of the deemed equity ratios as between Alberta and the

U.S., the Commission agrees with the following submission from Mr. Coyne:

With respect to the differences in equity ratios, this is explained in part by the process

U.S. regulators use for setting equity ratios versus their Canadian counterparts, where

equity ratios are deemed. As such, I have not recommended an adjustment for the

difference in equity ratios between the U.S. and Canada, as I believe the difference may

be justified by the use of deemed equity ratios in Canada versus greater reliance by U.S.

regulators on actual capital structures in comparison to peer companies.832

657. Mr. Coyne’s observation is supported by DBRS in its analysis of the regulatory

framework for utilities in Canada and the U.S. DBRS stated:

For some utilities, returns are based on the actual capital structure which is set within a

range determined by the state regulator. Pennsylvania is an example, where the

commission intervenes only if quarterly disclosed equity ratios fall outside a reasonable

range.833

658. In the 2009 GCOC decision, the Commission found that that the equity ratios in the U.S.

are likely higher as a result of the ability of management in certain U.S. jurisdictions to set the

capital structure within a range acceptable to the regulator. This is a differentiating point between

regulation of U.S. and Canadian utilities and an indication that allowed capital structures for U.S.

utilities should not be held up as representative of the capital structures required by Canadian

utilities in order to satisfy the fair return standard.834 The Commission continues to be of this

view.

659. Dr. Villadsen indicated that the average of the deemed equity ratios most recently

awarded in Canada is 36.59 per cent for electricity distributors and 39.86 per cent for natural gas

distributors. These figures include the deemed equity ratios awarded by the Commission in 2017

for the utilities in Alberta.835 In order to make a valid comparison between the deemed equity

ratios approved in Alberta and other Canadian jurisdictions, the Commission finds that the

deemed equity ratios for the utilities in Alberta must be excluded from the figures presented by

Dr. Villadsen. Omitting the deemed equity ratios awarded by this Commission, the averages for

the other Canadian jurisdictions are (1) 40 per cent for Canadian gas distributors; (2) 36.41 per

cent for Canadian electricity distributors; and (3) 35 per cent for Canadian electric transmission

companies.836

831 Exhibit 22570-X0913, paragraph 192. 832 Exhibit 22570-X0131, PDF page 87. 833 Exhibit 22570-X0842, PDF page 19. 834 Decision 2009-216, paragraph 193. 835 Exhibit 22570-X0766, PDF pages 49-50. 836 Exhibit 22570-X0766, PDF pages 49-50.



660. With regard to Dr. Villadsen’s comparison of the deemed equity ratios for Canadian gas

distributors, which average 40 per cent in other jurisdictions, the Commission observes that there

is a wide range of deemed equity ratios that make up this average, from 30 to 46.50 per cent. The

deemed equity ratios of 37 per cent and 39 per cent for ATCO Gas and AltaGas, respectively,

approved in Section 9.11 and Section 10, respectively of this decision are within this range.

Further, when these deemed equity ratios are compared to the approved deemed equity ratio of

36 per cent for the two large gas distribution utilities in Ontario, Union Gas Limited and

Enbridge Gas Distribution Inc., the deemed equity ratios of ATCO Gas and AltaGas are

higher.837 With regard to comparing the deemed equity ratio for electric distribution and

transmission utilities, the deemed equity ratio of 37 per cent approved in Section 9.11 of this

decision is 59 bps greater than the average for other Canadian electricity distributors, and 200

bps greater than the average for other Canadian electric transmission companies.

9.4 Industry financing practices

661. Mr. Hevert indicated that one of the focuses of his recommended deemed equity ratio

was on industry financing practices.838 He submitted that the capital structure of a utility must

support the financial strength of the utility during normal market conditions and during periods

of market uncertainty. Mr. Hevert described a key financing practice known as “maturity

matching” that applies when optimizing capital structure. The goal of maturity matching is to

align the average life of the securities in the capital structure with the average lives of the capital

assets being financed. He explained that the perpetual life of common equity mitigates

refinancing risk, whereas relying more heavily on debt increases the risk of refinancing maturing

debt obligations during less accommodating market environments. Mr. Hevert submitted that the

long-term nature of refinancing risks is not reflected in the near-term, pro forma credit metrics

used by the Commission to determine deemed equity ratios.839

662. Mr. Hevert stated that capital structure management is focused on multiple factors, some

that are company specific and others that are market-dependent. He indicated that these factors

are dynamic and complex, and are forward looking. Mr. Hevert suggested that utility capital

structure decisions recognize the long-term nature of the assets that support utility operations, the

need for short-term financial liquidity, and the fact that capital market conditions are not always

accommodating. He submitted that because utility operations are so dependent on capital market

access, his belief is that those considerations should extend to the factors that will enable the

cost-effective, efficient and timely access to capital over the long term under both constrained

and accommodating market conditions.840

Commission findings

663. Financial strength of the utilities is one of the factors the Commission considered as part

of its determination of the approved ROE and deemed equity ratios for 2018 to 2020. This is

evidenced by the Commission’s targeting of credit ratings in the A-range for the affected

utilities, as discussed in Section 9.6 and Section 9.7. An A-range credit rating should support the

financial strength of the utilities under varying market conditions, and help to ensure capital

attraction.

837 Exhibit 22570-X0766, PDF page 49. 838 Exhibit 22570-X0153.01, PDF page 7. 839 Exhibit 22570-X0153.01, PDF page 104. 840 Exhibit 22570-X0741.01, PDF page 75.



664. Mr. Hevert commented on the long-term nature of refinancing risks, but he did not

provide any type of detailed analysis that would help guide the Commission with respect to this

risk. Given the long-term nature of utility assets, and especially considering that the age of the

electricity transmission assets completed under the big build will be less than 10 years at the end

of 2020, the Commission considers that assessing long-term refinancing risk for the utilities in

Alberta will involve long-term assumptions about bond markets, which could change

substantially over the ensuing years.

665. The Commission observes that the affected utilities have issued a substantial amount of

30-year debt and some 40- and 50-year debt as well in recent years. All this debt matures as a

balloon payment at the end, except for ENMAX’s ACFA funding, which is a debenture by which

capital is paid off during the term. These debt issuances typically fund long-life assets of 40 to 50

years. At the end of 30 years, where a 40-year asset has been funded by 30-year debt,

approximately three-quarters of the asset’s cost will have been recovered already in depreciation

rates. Accordingly, the refinancing risk is only on one quarter of the original cost, which will

also be reduced by the equity component, and that is after 30 years of accumulated inflation so

any historic cost would not be significant, and the residual refinancing risk is only for a term of

10 years. The Commission has not been presented with any evidence that would indicate that this

risk is substantial.

666. Even if rates did increase dramatically, debt costs are a flow-through item to ratepayers

and the regulated firm can pass through these costs. In that way the utility is insulated from any

refinancing risk.

667. Mr. Hevert stated that the perpetual life of common equity mitigates against refinancing

risk. The Commission considers that unless the utility is financed 100 per cent by equity, there

will always be some level of refinancing risk; however, that risk is minimal and flowed through

to ratepayers.

9.5 Factors raised by FortisAlberta

668. FortisAlberta submitted that the determination of a deemed equity ratio for 2018 to 2020

should recognize the importance of equity funding in meeting the utility’s overall capital funding

requirements, and seek to foster the utility’s ability to attract capital from both equity and debt

investors on reasonable terms, and in required amounts. It contended that the need to foster

equity funding is heightened during the 2018 to 2020 period, because of the 2018 to 2022 PBR

term, and the provincial government’s climate leadership plan.841 FortisAlberta stated it will be

an important contributor to the development of distributed generation in Alberta, and it indicated

this will require significant amounts of capital investment. It submitted that the Commission’s

approval of an increase in the utility’s deemed equity ratio will help to ensure that the required

capital is available.842

669. FortisAlberta commented that the large amount of debt that currently comprises its

capital structure may not be sustainable as capital markets continue to evolve, and future debt

rates may not always be as low as they currently are. It submitted that these concerns with debt

841 Exhibit 22570-X0228, paragraph 4. 842 Exhibit 22570-X0228, paragraph 35.



should be recognized by the Commission and addressed by increasing the deemed equity ratio, in

order to attract more equity investment.843

670. Mr. Thygesen responded that the credit spreads for FortisAlberta have decreased almost

50 per cent since the time of the 2016 GCOC oral hearing, and its recent debt issue in September

2017 had a credit spread that was much lower than the spread on the debt issues in 2015 and

2016.844 He indicated that FortisAlberta does not face any refinancing risk until 2034.845

671. FortisAlberta cautioned that the reduction to its deemed equity ratio as a result of the

2016 GCOC decision potentially impairs its ability to attract equity in the future. It explained

that because it has one equity investor, and that equity investor has investments in other

regulated utilities, FortisAlberta’s ability to obtain future equity injections may be impaired to

the extent that it cannot demonstrate that it represents an investment possessing value,

comparable to the other regulated utilities owned by its investor.846

Commission findings

672. As discussed in Section 9.11, the Commission considers that the ROE and deemed equity

ratio it has approved for FortisAlberta as part of this decision satisfies the fair return standard.

The fair return standard includes consideration of the capital attraction, financial integrity and

comparability factors. The Commission considers this fosters FortisAlberta’s ability to attract

debt and equity capital.

673. FortisAlberta stated that the large amount of debt in its capital structure may not be

sustainable as capital markets evolve, but it provided no evidence about what this evolution may

entail. The submission from FortisAlberta that future debt rates may not always be as low as they

currently are, is likewise not supported by any evidence.

674. With respect to the potential impairment of FortisAlberta’s ability to attract equity in the

future, the Commission considers that FortisAlberta’s equity investor would likely assess the

approved ROE as well as the actual ROEs that have been achieved by FortisAlberta. The actual

ROEs achieved by FortisAlberta have averaged 10.2 per cent over the years 2010 to 2016.847

9.6 Maintaining credit ratings in the A-range

675. In the 2016 GCOC decision, the Commission noted its historical recognition of the

importance of maintaining a credit rating in the A-range. The Commission explained that an

objective of the analysis it undertook when establishing the approved deemed equity ratios was

to ensure that the deemed equity ratio, when combined with the approved ROE, would achieve

target credit ratings in the A-range.848

676. In this proceeding, the Commission asked Dr. Villadsen, Dr. Carpenter, Mr. Buttke,

Mr. Coyne, Mr. Thygesen, Mr. Madsen, Mr. Bell, Dr. Cleary and Mr. Johnson to comment on

843 Exhibit 22570-X0228, paragraphs 16-17. 844 Exhibit 22570-X0551, paragraphs 158-159. 845 Exhibit 22570-X0551, paragraph 160. 846 Exhibit 22570-X0228, paragraphs 19-20, 22. 847 Exhibit 22570-X0561.01, worksheet WP 4 – ROE Variation. 848 Decision 20622-D01-2016, paragraph 345.



whether it should continue to recognize the importance of maintaining a credit rating in the

A-range for the affected utilities.

677. Mr. Hevert stated that he considered the importance of maintaining an A-range credit

rating when he developed his recommended deemed equity ratio.849 Dr. Carpenter indicated his

agreement with the comments provided by Dr. Villadsen and Mr. Buttke.850

678. Mr. Hevert,851 Mr. Coyne,852 Dr. Villadsen853 and Mr. Buttke854 commented that having

A-range credit ratings helps ensure the affected utilities are able to attract capital from the long-

term Canadian bond market in almost all market conditions. Mr. Coyne noted that the BBB bond

market is not as well established in Canada as it is in the U.S., and consequently there is less

trading of sub-A rated debt in the Canadian credit market.855 Mr. Buttke suggested that a number

of large Canadian insurance companies may not buy long-term utility debt that is BBB rated.856

Mr. Buttke’s suggestion was supported by Mr. Coyne, who stated that life-insurance companies

and pension funds account for the bulk of the demand in long-term debt, and many of these

companies are limited to A-range rated securities.857

679. Mr. Buttke submitted that while credit ratings in the A-range are not guarantees of a fair

rate of return, they help reassure investors that some minimum level of return will likely be

forthcoming in this GCOC proceeding, as well as future GCOC proceedings.858

680. Dr. Villadsen stated that bond investors and DBRS have recognized the Commission’s

A-range credit rating policy as a sign of regulatory support for the utilities.859 Mr. Buttke

suggested that if the Commission renounced this policy, investors would change their

assumption about the level of regulatory support in Alberta.860 Dr. Villadsen commented that debt

investors and credit rating agencies would have a negative perception of any deviation in the

Commission’s policy.861

681. Dr. Villadsen stated that all else equal, sub-A rated debt has higher rates than debt rated

in the A-range.862 Mr. Madsen and Mr. Thygesen,863 as well as Dr. Cleary,864 provided

information on the current credit spreads for BBB and A-rated utilities, which suggested that the

differences were in the range of 18 bps to 30 bps. Mr. Coyne submitted that the difference in

credit spreads has narrowed at the current time because of the strong credit markets.865

849 Exhibit 22570-X0153.01, PDF page 109. 850 Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-001. 851 Exhibit 22570-X0153.01, PDF page 109. 852 Exhibit 22570-X0286, EPC-AUC2017NOV21-011. 853 Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-001. 854 Exhibit 22570-X0749, A16. 855 Exhibit 22570-X0775, PDF page 12. Transcript, Volume 6, page 1263. 856 Exhibit 22570-X0749, A16. 857 Exhibit 22570-X0775, PDF page 12. 858 Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-001. 859 Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-001. 860 Exhibit 22570-X0308, AUI/ATCO-AUC-2017NOV17-001. 861 Exhibit 22570-X0767.01, A111. 862 Exhibit 22570-X0308, AUI/ATCO-AUC2017NOV21-001. 863 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-001. 864 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005. 865 Exhibit 22570-X0775, PDF page 12.



Mr. Buttke commented that the magnitude of the credit spread differences would vary,

depending on market conditions.866 He pointed out that during the 2016 GCOC proceeding, the

Royal Bank of Canada estimated that the relative spread on 30-year debt between A-rated bonds

and BBB-rated bonds could have been nearer 100 bps.867

682. Mr. Thygesen, Mr. Madsen and Dr. Cleary focused on this difference in debt rates, as

well as the corresponding reduction in the deemed equity ratio, that they suggested could

accompany the targeting of credit ratings in the sub-A range.

683. Mr. Thygesen and Mr. Madsen submitted that the Commission should only target an

A-range credit rating if the higher interest costs being avoided by targeting the A-range are

greater than the higher cost of equity capital that results from maintaining the A-range credit

rating.868 Mr. Bell echoed these comments when he submitted that the lowest cost alternative to

provide safe and reliable utility service should be incorporated into customer rates.869

684. Mr. Johnson stated the Commission’s primary responsibility is to the fair return standard,

and suggested that an A-range credit rating should only be maintained without exceeding what is

required by the fair return standard.870

685. Dr. Cleary noted that there is a trade-off in deciding how much relief to provide the

utilities, in order to help them achieve a credit rating in the A-range.871 He suggested that if it

becomes extremely expensive to provide high equity ratios or an approved ROE to maintain an

A-range credit rating, it could make sense to allow one metric to slip a little bit into the triple B

plus range, rather than to strictly target credit metrics consistent with an A-range credit rating.872

Dr. Cleary submitted that the Commission should look at the circumstances of each utility, and

use judgment as well as analysis, including a cost-benefit analysis, to determine if it is too

expensive or infeasible to maintain an A-range credit rating.873

686. Mr. Madsen developed a quantitative model as a cost/benefit analysis.874 Based on the

outputs of his model, under different scenarios, Mr. Madsen submitted that a reduction in the

deemed equity ratios could be made, up to a certain limit. He cautioned that his submission is

dependent upon the credit spreads used as inputs to his model.875

687. Mr. Coyne contended that Mr. Madsen’s conclusions do not account for the additional

cost of equity that would be required due to the increased financial risk. Mr. Coyne stated that an

equity investor with an increase in financial risk because of a lower deemed equity ratio would

not be expected to have the same cost of equity as an investor with lower financial risk.876

866 Exhibit 22570-X0308, AUI/ATCO-AUC2017NOV21-001. 867 Exhibit 22570-X0749, A16. 868 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-001. 869 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005. 870 Exhibit 22570-X0667, CALGARY-AUC-2018JAN26-001. 871 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005. 872 Transcript, Volume 10, page 2055. 873 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005. 874 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-001. 875 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-001. 876 Exhibit 22570-X0775, PDF page 11.



688. Mr. Coyne submitted that the logic of finding the lowest possible credit rating such that

the marginal cost of debt does not exceed the cost of equity ignores the comparability

requirement of the fair return standard.877

Commission findings

689. After considering the evidence and submissions, the Commission is not prepared, at this

time, to depart from its historical practice of maintaining credit ratings in the A-range for the

affected utilities.

690. Mr. Madsen, Mr. Thygesen and Dr. Cleary focused their attention on the quantitative

aspect of the trade-off between the cost of increased ROE and deemed equity ratio needed to

maintain the lower debt cost associated with an A-range credit rating. The Commission considers

that while such a quantitative analysis is required, it is also necessary to consider qualitative

factors. The Commission agrees with Dr. Villadsen that this matter cannot be addressed simply

by considering the differences in debt rates between A-range debt and BBB-rated debt.

691. Mr. Madsen cautioned that the results of his quantitative analysis are dependent upon the

credit spreads used, which at this time are quite narrow as between A-rated debt and BBB-rated

debt. However, no evidence was presented as to whether this narrow difference is a long-term

trend. The Commission also agrees with Mr. Coyne that Mr. Madsen’s quantitative model does

not demonstrate whether he accounted for the additional cost of equity that might be required

due to lower deemed equity ratios being awarded.

692. The Commission finds that the qualitative factors put forward by Dr. Villadsen,

Mr. Hevert, Mr. Coyne and Mr. Buttke are valid considerations, and provide support for

targeting credit ratings in the A-range.

693. S&P’s current regulatory advantage assessment of Alberta is strong, but still on a

negative trend.878 The Commission considers it important to maintain this strong regulatory

assessment, and targeting the maintenance of credit ratings in the A-range plays an important

role in achieving this.

694. The use of the A-range credit rating target is a factor that respects the financial integrity,

capital attraction and comparability aspects of the fair return standard. Generally, utilities with

A-range credit ratings can obtain debt rates that are lower than utilities with sub-A rated debt.

Lower debt rates help bolster financial integrity. Credit ratings in the A-range help foster the

attraction of debt investors, as submitted by Mr. Hevert, Mr. Coyne, Dr. Villadsen and

Mr. Buttke. Regarding the attraction of equity capital, Mr. Buttke submitted that while credit

ratings in the A-range are not guarantees of a fair rate of return, they help reassure investors that

some minimum level of return will likely be forthcoming. The Commission considers that

maintaining credit ratings in the A-range, when combined with a sufficient ROE, meets the fair

return standard.

877 Exhibit 22570-X0775, PDF pages 11-12. 878 Exhibit 22570-X0141, paragraph 34.



695. Based on these findings, in combination with the ROE approved in Section 8.8 above, the

targeting of credit ratings in the A-range is one of the factors the Commission will continue to

use as part of its determination of the deemed equity ratios for 2018 to 2020.

9.7 Credit ratings and credit metric analysis

696. Dr. Villadsen,879 Mr. Hevert,880 Mr. Coyne,881 Mr. Madsen882 and Dr. Cleary883 each took

the position that their respective recommended deemed equity ratios either considered credit

metrics, or were supported by a credit metric analysis. In past GCOC decisions, the Commission

has placed weight on credit metrics.

9.7.1 Financial ratios, capital structure and actual credit ratings

697. Credit ratings assess the credit worthiness of a firm as determined by a credit rating

agency. A higher credit rating signals higher confidence in the firm’s ability to meet its interest

payments and to repay debt principal, allowing the company to borrow at a lower interest rate.

698. Credit metrics (or financial ratios) are an important, although not the only, component

that credit rating agencies consider when assessing the risk of any particular company and

assigning a credit rating. As noted in the 2016 GCOC decision, the Commission has historically

assessed three principal credit metrics:884

EBIT coverage: This is referred to as an interest coverage ratio. In the Commission’s

credit metric model, it is calculated by grossing up the net income by the statutory

income tax rate, adding the return on debt amount, and dividing the resulting figure by

the sum of the return on debt amount and the interest on the CWIP balance, calculated

using the deemed debt ratio and the embedded average debt rate.

FFO coverage: This is also an interest coverage ratio. In the Commission’s credit metric

model, it is calculated by adding the return on debt amount, the net income and the

depreciation collected and dividing the resulting figure by the sum of the return on debt

amount and the interest on the CWIP balance, calculated using the deemed debt ratio and

the embedded average debt rate. It is important to note that in the Commission’s credit

model, the interest expense associated with the CWIP balance is not included in the

numerator because it is based on the assumption that there is no CWIP included in rate

base.

FFO/debt: S&P compares this payback ratio against benchmarks to derive the

preliminary cash flow/leverage assessment for a company. S&P notes that this ratio is

also useful in determining the relative ranking of the financial risk of companies.885 In the

Commission’s credit metric model, it is calculated by adding the net income and the

879 Exhibit 22570-X0193.01, A5. 880 Exhibit 22570-X0153.01, PDF page 123. 881 Exhibit 22570-X0131, PDF pages 100-101. 882 Exhibit 22570-X0557, paragraph 118. 883 Exhibit 22570-X0562.01, PDF page 6. 884 Decision 20622-D01-2016, paragraph 356. 885 Exhibit 20622-X0089, PDF page 736.



depreciation collected and dividing the resulting figure by the sum of the deemed mid-

year debt for rate base and CWIP.

699. In the 2016 GCOC decision, the Commission took guidance from the EBIT coverage

ratio threshold used in the 2009 GCOC proceeding,886 in which the Commission observed that an

EBIT coverage of 2.0 was the minimum threshold associated with regulated utilities with an

A-range credit rating.

700. In the 2016 GCOC decision, the Commission also placed greater weight on S&P’s credit

metric benchmarks for FFO coverage and FFO/debt, using a “low volatility scale.” The

Commission noted that the credit metric benchmarks used by S&P for an A-range credit rating

are an FFO coverage ratio of 2.0 to 3.0, an FFO/debt ratio of 9.0 per cent to 13.0 per cent, and an

EBITDA coverage ratio of 2.5 to 4.0. The Commission did not focus on the EBITDA coverage

ratio in the 2016 GCOC decision.887

701. In the 2016 GCOC decision, the Commission also calculated the deemed equity ratios

that were required to attain the minimum credit metrics necessary to maintain an A-range credit

rating for a typical taxable distribution utility, a typical non-taxable distribution utility, a typical

taxable transmission utility and a typical non-taxable transmission utility. The Commission has

performed the same calculations as part of this decision.

Mr. Hevert’s comments on credit metrics

702. Mr. Hevert stated that while credit metrics are important inputs into the credit rating

process, they are only one consideration used by the credit-rating agencies.888 He explained that

in assessing credit ratings, DBRS and S&P consider many factors, including the quality of the

regulatory regime, as well as historical and forward-looking credit ratios.889

703. Mr. Hevert noted that there is considerable variation in the historical Rule 005 data that is

used by the Commission and, as a result, it is not certain that the parameters used in the

Commission’s credit metric calculations will equal those likely to be observed in 2018, 2019 and

2020.890

704. Mr. Hevert demonstrated how changes in two parameters (mid-year CWIP percentage

and depreciation) would have affected the FFO/debt credit metric calculated by the Commission

in the 2016 GCOC decision. He indicated that if the actual CWIP percentage is higher than

expected, this will reduce actual cash flows and the FFO/debt ratio. If the actual depreciation

parameter is lower than expected, this will also reduce actual cash flows and the FFO/debt ratio.

Mr. Hevert pointed out that the Commission should understand how these variations affect the

credit metrics.891

705. Mr. Hevert indicated that when S&P calculates ratios using debt, it makes several

adjustments to increase debt balances to reflect debt-like financial obligations. He suggested that

886 Decision 20622-D01-2016, paragraphs 357-358, 399. 887 Decision 20622-D01-2016, paragraphs 393, 399. 888 Exhibit 22570-X0153.01, PDF page 111. 889 Exhibit 22570-X0153.01, PDF page 121. 890 Exhibit 22570-X0153.01, PDF pages 111-112. 891 Exhibit 22570-X0153.01, PDF pages 114-119.



if these adjustments are not reflected in the Commission’s credit metric calculations, then the

FFO/debt ratio would be overstated.892 He noted that S&P focuses on variability in earnings when

it assesses future expected credit quality, while the only source of variation in the Commission’s

credit metric calculations is the deemed equity ratio.893

Mr. Coyne’s comments on credit metrics

706. Mr. Coyne indicated that credit metrics are one consideration in assessing financial

risk.894 He acknowledged the three credit metrics used by the Commission in the 2016 GCOC

decision, and the A-range thresholds established for these metrics by S&P. He stated that another

core ratio used by S&P is the debt/EBITDA ratio, and that the A-range threshold used by S&P

for this ratio is 4.0 to 5.0.895

707. Mr. Coyne calculated the following ratios as part of his credit metric analysis (1) EBIT

coverage; (2) EBITDA coverage; (3) FFO coverage; (4) FFO/debt; and (5) debt/EBITDA.896

708. Mr. Coyne calculated and reported these credit metrics as of December 31, 2016, for

(1) each of the companies in his U.S. electric proxy group; (2) each of the companies in his

Canadian utility proxy group; (3) each of the companies in his North American electric proxy

group; (4) the transmission utilities in Alberta; (5) the non-taxable transmission utilities in

Alberta; (6) the distribution utilities in Alberta; and (7) the non-taxable distribution utilities in

Alberta.897

709. Based on the credit metrics he calculated and reported, Mr. Coyne submitted that the

utilities in Alberta are well below the median for his North American utility proxy group, and in

the majority of cases, are very near the bottom, which indicates a financially vulnerable risk

profile. He added that all of the interest coverage ratios for the affected utilities are very low,

when compared to his North American utility proxy group. Mr. Coyne submitted that this

additional financial risk needs to be considered, in combination with the elevated business risks,

to determine an appropriate level of equity for the affected utilities.898

710. Mr. Coyne pointed out that, based on his credit metric calculations, the non-taxable

distribution and transmission utilities in Alberta would fall below S&P’s A-range thresholds for

the EBIT coverage ratio and the debt/EBITDA ratio, and the transmission utilities would fall

below the debt/EBITDA ratio. He stated that the deemed equity ratios approved in 2016 do not

meet the Commission’s objective of satisfying the credit metric requirements for an A-range

credit rating.899

892 Exhibit 22570-X0153.01, PDF page 113. 893 Exhibit 22570-X0153.01, PDF pages 113-114. 894 Exhibit 22570-X0131, PDF page 89. 895 Exhibit 22570-X0131, PDF page 91. 896 Exhibit 22570-X0131, PDF pages 92-93. 897 Exhibit 22570-X0131, PDF pages 91-92. 898 Exhibit 22570-X0131, PDF page 94. 899 Exhibit 22570-X0131, PDF pages 98-99.



Mr. Thygesen’s comments on credit metrics

711. Mr. Thygesen submitted that any credit metric calculations for EPCOR should use the

ACFA debt rate.900 He suggested that the reason why EPCOR’s embedded average debt rates are

greater than an average Alberta utility is because since 2013, the credit spreads for Westcoast

Energy Inc., one of the four comparator companies EPCOR uses to establish its stand-alone debt

rates, are much higher than the other three companies in the comparator group. Mr. Thygesen

suggested that Westcoast Energy Inc. has little in common with EPCOR, and EPCOR should not

include this company as part of its comparator group. He recommended that any credit metric

calculations for EPCOR as part of this GCOC proceeding reflect the exclusion of Westcoast

Energy from the comparator group.901

EPCOR’s comments on credit metrics

712. EPCOR reiterated its position that any issue with respect to its debt rates is beyond the

scope of this GCOC proceeding. It stated it continues to address the issues raised by

Mr. Thygesen in its tariff-related proceedings. EPCOR added it has placed evidence on the

record that completely refutes Mr. Thygesen’s claim about the use of Westcoast Energy skewing

the credit spread information, which EPCOR submitted was ignored by Mr. Thygesen.902 EPCOR

pointed out that debt rates do not impact the FFO/debt credit metric, which is the metric that

Mr. Madsen places the majority of weight on in his credit metric analysis.903

713. EPCOR noted that the pro forma credit metrics for its distribution and transmission

functions are generally lower than those the Commission calculates for the generic distribution

and transmission utilities. EPCOR stated the differences are because it has a higher than average

embedded debt rate, a lower than average depreciation rate, a lower than average CWIP

percentage, and because it is income tax exempt. It noted that DBRS foresees a deterioration in

the cash-flow to debt ratios for EPCOR’s distribution and transmission functions, and DBRS

expects these ratios to decrease to the BBB-rating range.904

Mr. Bell’s comments on credit metrics

714. Mr. Bell’s submitted that his base case credit metrics show that an increase in the deemed

equity ratio is not required. He indicated that a deemed equity ratio of 35 per cent would satisfy

the EBITDA coverage, FFO interest coverage and FFO/debt targets for an A-range credit rating

established by S&P, and a slightly higher deemed equity ratio would satisfy the thresholds

established by DBRS.905

Mr. Madsen’s comments on credit metrics

715. Mr. Madsen agreed with the Commission’s increased reliance, as part of the 2016 GCOC

decision, on the S&P credit metrics and related thresholds necessary for an A-range credit

rating.906

900 Exhibit 22570-X0551, paragraph 19. 901 Exhibit 22570-X0551, paragraphs 20-27. 902 Exhibit 22570-X0733, A45. 903 Exhibit 22570-X0733, A43. 904 Exhibit 22570-X0195, paragraphs 67, 70, 74, 77. 905 Exhibit 22570-X0559, A17. 906 Exhibit 22570-X0557, paragraph 165.



716. Mr. Madsen stated that forecast capital expenditures for the Alberta utilities in 2018,

2019 and 2020 are lower than historical levels, which will also result in lower debt issuances

over this period. He commented that because of this, there will be less pressure on credit metrics.

Mr. Madsen submitted that this reduces business risk, and supports his conclusion for lower

deemed equity ratios over the test period.907

717. As part of Mr. Madsen’s credit metric analysis, he estimated the weighting that the

Commission applied to the three credit metrics it used in the 2016 GCOC decision.908

Mr. Madsen indicated that the input parameters he used in his credit metric analysis were

calculated using information from the Rule 005 filings the utilities submitted in 2017, which

reported on the results for 2016. He used a simple average for debt rates, and a weighted average

for depreciation and CWIP. Mr. Madsen submitted that the most current information from a

single year should be used for the input parameters, rather than a number of historical years,

because the deemed equity ratio is being approved on a forward basis.909

718. Based on his credit metric calculations, and his estimation of the weighting the

Commission placed on the three credit metrics in the 2016 GCOC decision, Mr. Madsen

concluded that a base level equity ratio of 35.8 per cent would allow the transmission utilities to

achieve an A-range credit rating, and a base level equity ratio of 36.5 per cent would allow the

distribution utilities to achieve an A-range credit rating.910

719. Turning to his assessment that business risk for the transmission utilities has decreased

since the 2016 GCOC decision into account, Mr. Madsen stated that the deemed equity ratio for

the Alberta transmission utilities should be set at a base level of 35.5 per cent. Combining his

assessment of a decline in the business risk of the distribution utilities since the 2016 GCOC

decision, with their higher overall credit metric levels, Mr. Madsen stated that the deemed equity

ratio for the Alberta distribution utilities should be set at a base level of 36 per cent. He noted

that neither of these recommended base levels reflect the possible adoption of the future income

tax method.911

Dr. Villadsen’s comments on credit metrics

720. Dr. Villadsen recommended that the Commission select a capital structure that is

sufficient to meet credit metric thresholds toward the middle of the published guidelines of all

the major credit-rating agencies, including DBRS, Fitch Ratings, Moody’s Investor Services

(Moody’s) and S&P.912

721. Dr. Villadsen suggested that even without a credit-rating downgrade, operating with

credit metrics at the low end of the scale could place AltaGas and the ATCO Utilities at risk of

decreased access to credit, or higher debt rates, in the event of an unexpected financial downturn

in the economy.913

907 Exhibit 22570-X0557, paragraphs 220 and 222. 908 Exhibit 22570-X0557, paragraph 243. 909 Exhibit 22570-X0557, paragraphs 244-249. 910 Exhibit 22570-X0557, paragraphs 253 and 261. 911 Exhibit 22570-X0557, paragraphs 262-264. 912 Exhibit 22570-X0193.01, A85, A90. 913 Exhibit 22570-X0193.01, A85.



722. Dr. Villadsen was concerned with the Commission’s reliance in the 2016 GCOC decision

on S&P’s credit metric thresholds using the low volatility table. She indicated that S&P only

applies the low volatility table if it assesses the utility’s regulatory advantage score as being

strong. While she acknowledged that S&P’s regulatory advantage score for Alberta is strong, she

noted it is with a negative trend.914 Dr. Villadsen submitted that, based on Dr. Carpenter’s

evidence, the business risk and regulatory risk for AltaGas and the ATCO Utilities is increasing,

which reduces their ability to fit into S&P’s low volatility category. She also noted the concerns

expressed by credit rating agencies about the quality of regulatory support in Alberta.

Consequently, Dr. Villadsen cautioned that it is risky to assume that S&P, as well as other rating

agencies, will continue to evaluate credit metrics under the assumption of a strong regulatory

environment in Alberta.915

723. Noting that the Commission placed greater weight on S&P’s credit metric guidelines in

the 2016 GCOC decision, but considering her concerns with only relying on the low end of the

credit metric guidelines established under S&P’s low volatility table, Dr. Villadsen submitted

that it is more appropriate for the Commission to target the two FFO-based credit metrics using

the point of overlap between S&P’s low volatility table and medial volatility table. This would

set the FFO interest coverage ratio threshold at 3.0, and the FFO/debt ratio threshold at 13.0.916

724. Dr. Villadsen noted that in Mr. Bell’s credit metric calculations, he incorrectly assumed

an income tax rate of 27 per cent when he calculated EBIT and EBITDA.917 She noted that

Mr. Madsen did the same when he calculated EBIT.918 She stated that this is greater than the

effective income tax rate that AltaGas and the ATCO Utilities will have under the flow-through

income tax method, and consequently, the EBIT and EBITDA figures are too high.919 While she

acknowledged the Commission’s determinations in the 2011 GCOC decision about using the

statutory income tax rates to calculate EBIT, Dr. Villadsen submitted that it is proper to consider

credit metrics in a manner that is consistent with the income tax method that will be used during

the test period.920 She stated that calculating EBIT and EBITDA using the effective tax rates has

a substantial impact.921

725. Dr. Villadsen noted that Mr. Bell did not calculate the debt/EBITDA ratio.922 She

calculated the debt/EBITDA ratio using Mr. Bell’s inputs, and advised that it would require a

deemed equity ratio of 41 per cent in order to meet the minimum standard of S&P.923

726. Dr. Villadsen disagreed with Mr. Bell’s conclusion that based on his credit metric

calculation, a deemed equity ratio slightly higher than 35 per cent would satisfy the DBRS

thresholds for an A-range credit rating. She submitted that the required equity ratio would have

914 Exhibit 22570-X193.01, A88. 915 Exhibit 22570-X193.01, A89. 916 Exhibit 22570-X0193.01, A90. 917 Exhibit 22570-X0767.01, A105. 918 Exhibit 22570-X0767.01, A107. 919 Exhibit 22570-X0767.01, A105. 920 Exhibit 22570-X0767.01, A106. 921 Exhibit 22570-X0767.01, A108. 922 Exhibit 22570-X0767.01, A101-A102. 923 Exhibit 22570-X0767.01, A104.



to be at least 39 per cent, and this would only meet the low end of the DBRS threshold for

FFO/debt.924

727. Dr. Villadsen stated that Mr. Madsen’s approach of attempting to infer the weightings the

Commission assigned to the three credit metrics it used in the 2016 GCOC decision was

confusing and arbitrary, and yields nonsensical results. Dr. Villadsen pointed out that

Mr. Madsen ignored his calculated credit metric results when he recommended deemed equity

ratios.925

AltaLink’s comments on credit metrics

728. AltaLink suggested that because its business risk is now higher than it was during the

2016 GCOC proceeding, the obvious conclusion is that the credit metric ratio thresholds for the

2018 GCOC proceeding should be higher, in order to account for this increased risk.926 AltaLink

contended that an FFO/debt ratio below the 12.5 per cent absolute minimum of DBRS places it

at undue risk and is further evidence that a fair return has not been awarded.927

729. AltaLink noted that S&P has not removed the negative trend rating from its regulatory

advantage assessment of Alberta that was present during the 2016 GCOC proceeding. Given this

negative trend, and the increased uncertainties in its business risk since the 2016 GCOC

proceeding, AltaLink submitted that its FFO/debt ratio should be established “in a comfortable

range,”928 in order for it to be able to absorb the financial implications of the business risk

uncertainties it faces.929

730. AltaLink disagreed with the arbitrary weighting calculations that Mr. Madsen derived for

the FFO/debt, FFO interest coverage and EBIT interest coverage ratios. It submitted that the

deemed equity ratio cannot be derived from a formula that is based on an unreasonably low

FFO/debt ratio.930

731. AltaLink commented that the data used by Mr. Madsen and Mr. Bell to derive the inputs

for their credit metric calculations was from 2016 and outdated. It submitted that the most

current data available, including forecast data, should be used.931

732. AltaLink and EPCOR critiqued the credit metric calculations of Mr. Bell. They noted that

Mr. Bell used a weighted average for debt, whereas the Commission uses a simple average. They

noted that Mr. Bell used a simple average for depreciation, as opposed to the Commission’s use

of a weighted average. They commented that Mr. Bell did not perform separate calculations for

the transmission and distribution utilities, but instead, a combined calculation.932

924 Exhibit 22570-X0767.01, A103. 925 Exhibit 22570-X0767.01, A109. 926 Exhibit 22570-X0141, paragraph 32. 927 Exhibit 22570-X0141, paragraphs 37-38. 928 Exhibit 22570-X0141, paragraph 35. 929 Exhibit 22570-X0141, paragraphs 34-35. 930 Exhibit 22570-X0738, paragraphs 20-21. 931 Exhibit 22570-X0738, paragraph 33. Exhibit 22570-X0738, paragraph 48. 932 Exhibit 22570-X0738, paragraphs 44-46. Exhibit 22570-X0733, A34.



The UCA’s comments on credit metrics

733. The UCA opposed Dr. Villadsen’s recommendation that the Commission rely on the

medial volatility table established by S&P to assess credit metric thresholds. It submitted that

Dr. Villadsen has provided no evidence to support her claim that the credit rating agencies no

longer view Alberta as possessing a strong regulatory advantage. The UCA recommended that

the Commission apply the same credit metric thresholds that were applied in the 2016 GCOC

decision.933

The CCA’s comments on credit metrics

734. The CCA disagreed with AltaLink’s suggestion that forecast values be used to determine

the inputs used in a credit metric analysis. It submitted that the use of forecast data adds

significant new uncertainty into the credit metric process. The CCA suggested that if forecast

data is used, it be should be used uniformly across all the affected utilities.934

735. The CCA submitted that the Commission should not factor in the impacts of ACFA

funding on a utility’s credit metrics in determining the deemed equity ratio in order to maintain

a utility’s financial integrity.935

Commission findings

736. The Commission will, consistent with its approach in past GCOC decisions, and its

findings in Section 9.6, award deemed equity ratios that are, on a stand-alone basis, consistent

with credit ratings in the A-range.

737. In this proceeding, parties provided evidence regarding the benchmarks associated with

certain credit metrics used by various credit-rating agencies. The Commission acknowledges the

submission of Mr. Hevert that credit metrics are only one part of the credit-rating determination

process. However, the Commission notes that Mr. Hevert, as well as Dr. Villadsen, Mr. Coyne,

Mr. Bell and Mr. Madsen, assessed credit metrics as part of the analysis to determine

recommended equity ratios. The Commission is likewise satisfied that formal credit metrics

should be considered in the assessment of deemed equity ratios. In doing so, the Commission is

cognizant that the process of setting credit metrics required to maintain an A-range credit rating

for the utilities in Alberta is a function of market dynamics and credit agency analysis of macro-

economic trends, Canadian utility industry specific variables and future investor expectations,

applied to an assessment of the relative risk of the utility sector, and perceptions of the regulatory

environment.

738. Credit metrics reflect past market expectations as well as anticipated market expectations,

given an assessment of current economic conditions, the information and assumptions employed

in conducting the analysis and judgment of relative risk. The element of judgment is reflected to

some degree, in the differing credit metrics employed and the breadth of ranges used by various

credit rating agencies and market analysts. Further, the application of utility sector credit metrics

to a particular Alberta utility involves a further element of judgment on factors such as the

Alberta regulatory climate.

933 Exhibit 22570-X0897.01, paragraphs 245-246. 934 Exhibit 22570-X0888, paragraph 346. 935 Exhibit 22570-X0888, paragraph 385.



739. From a practical perspective, however, credit metrics affect investor risk perceptions and

consequently may affect market behaviour. The Commission considers the credit metrics

reflected in credit rating and market analyst reports to be generally reflective of future

expectations of utility debt and equity investors with respect to credit metric fundamentals. This

observation is supported generally by a review of actual market behaviour. The Commission

finds that, generally, most utilities in Alberta have had little difficulty raising debt and equity

financing on satisfactory terms while maintaining an A-range credit rating.

740. In the 2016 GCOC decision, the Commission placed greater weight on S&P’s credit

metric benchmarks for FFO coverage and FFO/debt, using a “low volatility scale.” No evidence

was submitted that this low volatility scale is no longer applicable for the utilities in Alberta. The

Alberta regulatory advantage is currently rated by S&P as “strong” with a trend of “negative.”936

This is the same rating that was in place during the 2016 GCOC proceeding. Further support for

the continued use of S&P’s low volatility scale is the fact that, in Section 9.3 of this decision, the

Commission found no significant increase in generic business risk for the affected utilities since

the 2016 GCOC proceeding.

741. Dr. Villadsen submitted that the Commission establish a capital structure that is sufficient

to meet the credit metric thresholds at the middle of the published guidelines of all the major

credit-rating agencies. However, she did not provide the thresholds that are used by Fitch

Ratings, and the Commission acknowledged in the 2016 GCOC decision that it is difficult to see

how any of the major regulated utilities in Canada could qualify for a credit rating of A from

Moody’s.937 The Commission continues to hold this view, especially with respect to the

FFO/debt ratio benchmark range of 18 to 26 per cent utilized by Moody’s. Consequently, the

Commission will not place any reliance on the benchmark ranges of Moody’s, nor will the

Commission consider Fitch Ratings in its assessment of the affected utilities’ credit metrics.

742. The DBRS benchmark ranges for equity ratio, EBIT coverage and FFO coverage were

examined during the 2016 GCOC proceeding. In that proceeding, the Commission found

evidence that cast doubt on the use of the credit metric benchmark ranges established by DBRS

to qualify for an A-range credit rating.938 No evidence was provided in this GCOC proceeding to

satisfactorily eliminate this doubt. Accordingly, the Commission finds the credit metric

benchmarks used by DBRS to be less informative than the S&P benchmarks in evaluating the

financial parameters necessary for an A-range credit rating.

743. Dr. Villadsen recommended that the Commission target the two FFO-based credit metrics

using the point of overlap between S&P’s low volatility and medial volatility tables, which

would set the FFO interest coverage threshold at 3.0, and the FFO/debt ratio threshold at 13.0.

AltaLink contended that a minimum FFO/debt ratio should be 12.5 per cent. The Commission

notes that the credit metric analysis it has subsequently prepared, using the approved ROE and

deemed equity ratio, shows FFO interest coverage ratios of 3.9 and 3.3 for the distribution and

transmission utilities, respectively, and FFO/debt ratios of 13.8 and 11.1 per cent for the

distribution and transmission utilities, respectively. Three of these four metrics exceed the

thresholds that Dr. Villadsen has recommended.

936 Exhibit 22570-X0188.01, PDF page 285. 937 Decision 20622-D01-2016, paragraph 395. 938 Decision 20622-D01-2016, paragraph 394.



744. In addition to the three credit metrics the Commission examined in the 2016 GCOC

decision, Mr. Coyne calculated the EBITDA coverage and debt/EBITDA ratios in this

proceeding. He noted that debt/EBITDA is a core ratio used by S&P, and the benchmark for this

ratio, using the low volatility table, is four to five. The Commission calculated the debt/EBITDA

ratios resulting from the 2016 GCOC decision credit metric model, and found that while the

taxable distribution utilities would have just reached the threshold of five, the non-taxable

distribution utilities and both the taxable and non-taxable transmission utilities would not have

reached the threshold. This situation remains in the current GCOC proceeding. Considering that

the utilities have been able to maintain credit ratings from S&P in the A-range without meeting

this credit metric threshold, the Commission considers that meeting the debt/EBITDA ratio is not

important, in and of itself. Consequently, the Commission will not focus its attention on the

debt/EBITDA credit metric.

745. With respect to the EBITDA coverage ratio, the Commission calculated the results of this

ratio resulting from the 2016 GCOC decision credit metric model, and found that both the

taxable and non-taxable distribution utilities, as well as the taxable and non-taxable transmission

utilities, would have exceeded S&P’s medial volatility table benchmark of 2.75. The four ratios

were all in excess of 3.1. The same situation remains in the current GCOC proceeding.

746. Mr. Coyne submitted that the credit metrics for the utilities in Alberta are very near the

bottom, when compared to the results of his North American electric proxy group. The

Commission considers this is mainly because of the higher approved ROEs and deemed equity

ratios that the U.S. utilities are awarded. The Commission has discussed the comparability of the

U.S. and Canadian regulatory regimes, deemed equity ratios and approved ROEs, in

Section 9.3.3, Section 9.3.4 and Section 8.1.

747. Mr. Hevert indicated that changes in the CWIP percentages and the depreciation

parameters used in the Commission’s credit metric model will affect the FFO/debt ratios. The

Commission is aware that changes in these parameters would have an effect on the FFO/debt

ratio, and this is one of the reasons the Commission does not target the FFO/debt ratio at the

lower end of the threshold. Even if the CWIP and depreciation parameters for the distribution

utilities were changed to 10 per cent and five per cent, respectively, the resulting FFO/debt ratio

would be 11.8 per cent, which is toward the upper end of S&P’s range. If the CWIP and

depreciation parameters for the transmission utilities were changed to 10 per cent and 3.8 per

cent, respectively, the resulting FFO/debt ratio would be 10 per cent, which is still in excess of

the nine per cent threshold established by S&P.

748. Mr. Hevert indicated that S&P makes several adjustments to increase debt balances to

reflect debt-like financial obligations, and if these are not reflected in the Commission’s credit

metric calculations, then the resulting FFO/debt ratio would be overstated. This issue was

addressed by the Commission in the 2016 GCOC decision, and the Commission determined that

any overstatement is not material.939 No evidence was brought forward in this proceeding to

support a contrary finding.

749. AltaLink contended that an FFO/debt ratio below 12.5 per cent places it at undue risk,

and is further evidence that a fair return has not been awarded. The Commission notes that the

939 Decision 20622-D01-2016, paragraphs 391-392.



12.5 per cent threshold for FFO/debt put forward by AltaLink is from DBRS. The Commission

has previously commented that the benchmarks established by DBRS are not as informative as

those used by S&P.

750. AltaLink also recommended that the Commission use the most current data available,

including forecast data, in calculating credit metrics. The Commission agrees with the CCA that

the use of forecast data adds uncertainty into the credit metric process. This is evidenced by the

difference in the 2018 forecast CWIP, debt cost and depreciation percentage parameters used by

AltaLink in the credit metric calculations included as part of its rebuttal evidence,940 and those it

submitted during the oral hearing.941 If forecast data was used for AltaLink, the Commission

would have to be consistent and use forecast data for all the affected utilities. Forecast data for

all the affected utilities, prepared at the same time and for the same years, was not provided. For

these reasons, the Commission will continue to base its credit metric analysis on actual data

provided through the Rule 005 reports.

751. Dr. Villadsen submitted that it is proper to use the effective income tax rates of the

affected utilities, instead of the statutory income tax rates, as part of the credit metric analysis.

She indicated that the average effective income tax rate for 2016 for the five utilities that paid

income taxes was 8.8 per cent.942

752. The Commission notes that the income tax rate does not have any effect on the FFO

ratios. In addition, if the Commission uses an income tax rate of 8.8 per cent in its credit metric

model, the resulting ratios for the distribution utilities would be an EBIT coverage of 2.1 and

EBITDA coverage of 4.0, which are within the Commission’s thresholds for an A-range credit

rating. Likewise, if an income tax rate of 8.8 per cent for the transmission utilities is used in the

Commission’s credit metric model, the resulting ratios would be an EBIT coverage of 2.1 and

EBITDA coverage of 3.4, which are within the Commission’s thresholds for an A-range credit

rating. The Commission will continue to analyze credit metrics using an income tax rate of

27 per cent and an income tax rate of zero. Effective income tax rates that are less than the

statutory income tax rates will fall somewhere in the range of the results for these two analyses.

753. Mr. Thygesen commented that any credit metric calculations for EPCOR should use the

ACFA debt rate, and it should also reflect the exclusion of Westcoast Energy from the

comparator group. ACFA debt rates are reflected in the Commission’s credit metric analysis,

because the average embedded debt rate includes the debt of ENMAX, which is ACFA debt. The

Commission will not address the use of Westcoast Energy as a comparator in establishing

EPCOR’s debt rates for the reasons addressed in Section 7.2, specifically that the Commission is

not approving debt rates for EPCOR in this proceeding. This issue is best addressed in a GTA or

rebasing proceeding.

754. EPCOR noted that its credit metrics are generally lower than the credit metrics calculated

by the Commission for an average distribution and transmission utility. EPCOR stated this is

because of its higher debt rates and lower depreciation rates. The Commission has previously

stated in Section 7.2 that the City of Edmonton’s refusal to make ACFA funding available to

EPCOR is reflected in EPCOR’s lower credit metrics. The Commission considers that EPCOR’s

940 Exhibit 22570-X0738, Table 3. 941 Exhibit 22570-X0858, Table 3. 942 Exhibit 22570-X0767.01, footnote 196.



use of the direct life method for depreciation, and the resulting treatment of salvage, is a

contributor to its lower than average depreciation rates. In the decision on EPCOR’s 2015 to

2017 transmission GTA, the Commission directed EPCOR to conduct and file research

respecting alternative methods of accounting for the cost of removal of retired assets.943 The

Commission will not provide additional credit metric relief to EPCOR on this basis.

755. Mr. Madsen determined base level equity ratios from his credit metric calculations and

his estimation of the weighting the Commission placed on the EBIT coverage, FFO coverage and

FFO/debt ratios in the 2016 GCOC decision. The Commission agrees with the submissions of

Dr. Villadsen and AltaLink that Mr. Madsen’s attempt to infer the weightings the Commission

placed on these three credit metric ratios in the 2016 GCOC decision was arbitrary, and yielded

results that were illogical.

756. Mr. Madsen inferred that for the distribution utilities in the 2016 GCOC decision, the

Commission weighted the FFO/debt ratio at 425 per cent, and the EBIT coverage and FFO

coverage ratios at negative 163 per cent.944 There is nothing in the 2016 GCOC decision that

would suggest the Commission used any numerical weightings for the three credit ratios.

757. The use of these weightings in determining his deemed equity ratio recommendations

leads to results that are not logical. Mr. Madsen’s credit metric calculations for ENMAX

Distribution showed that in order for it to achieve an A-range credit rating, it required a 39.6 per

cent deemed equity ratio. His calculations showed that EPCOR Distribution would require a

30.6 per cent deemed equity ratio in order to achieve an A-range credit rating. This is despite

Mr. Madsen’s submission that EPCOR Distribution has the worst credit metrics of any utility in

Alberta.

758. The Commission notes that, as set out in Table 8, for 2016 ENMAX Distribution’s debt

cost and CWIP percentages were significantly lower than those of EPCOR Distribution.

ENMAX Distribution’s depreciation percentage for 2016 was greater than that of EPCOR

Distribution. These three differences result in the credit metrics of ENMAX Distribution being

better than the credit metrics of EPCOR Distribution. However, based on the use of his inferred

weighting results, Mr. Madsen calculated that EPCOR Distribution only required a deemed

equity ratio of 30.6 per cent, whereas ENMAX Distribution required a deemed equity ratio of

39.6 per cent. These results are evidence that Mr. Madsen’s inferred weightings are not a proper

basis upon which to determine a recommended equity ratio.

759. For all the above reasons, the Commission did not find the methodology used by

Mr. Madsen to determine his recommended deemed equity ratios helpful, and the Commission

has assigned no weight to his deemed equity recommendations.

9.7.2 Equity ratios associated with credit metrics

760. In the 2016 GCOC decision (tables 20-23), the Commission provided a sensitivity

analysis to illustrate the effect of a range of equity ratios on the three principal credit metrics for

the distribution utilities and the transmission utilities, using income tax rates of 27 per cent and

943 Decision 3539-D01-2015: EPCOR Distribution & Transmission Inc., 2015-2017 Transmission Facility Owner

Tariff, Proceeding 3539, Application 1611027-1, October 21, 2015, paragraph 852. 944 Exhibit 22570-X0557, Table 11.



zero. The analysis was based on certain input parameters associated with the affected utilities.

The Commission has prepared a similar analysis as part of this decision.

761. The parameter values used by the Commission in the 2016 GCOC decision, as well as the

parameter values the Commission has decided to use in this proceeding, are set out in Table 8

below. The Commission’s reasons for selecting the updated parameter values follow.

Table 8. Parameters for calculating credit metrics

Parameter

Parameter values applied in 2016

GCOC decision – taxable

distribution utilities

Parameter values applied in 2016

GCOC decision – taxable

transmission utilities

Parameter values applied in this

decision – taxable distribution

utilities

Parameter values applied in this

decision – taxable transmission

utilities

%

Embedded average debt rate 4.80 4.80 4.70 4.70

ROE 8.30 8.30 8.50 8.50

Income tax rate 27.00 27.00 27.00 27.00

Depreciation 5.75 4.10 5.85 4.20

Construction work in progress 3.78 5.00 3.21 5.00

762. In arriving at the updated parameters, the Commission has reviewed the actual

parameters from the 2014 and 2015 Rule 005 filings set out in the 2016 GCOC decision, and the

2016 Rule 005 filings that were submitted as part of this proceeding.

763. The ROE input parameter is common to all utilities, as is the income tax rate input

parameter for those utilities that are not income tax exempt. The Commission has summarized

the embedded average debt rates, depreciation rates and CWIP percentages for each affected

utility in Table 9.



Table 9. Embedded average debt rates, depreciation rates and CWIP percentages by utility

Utility Invested capital

($000) Debt cost

%

Depreciation as a percentage of

invested capital

Mid-year CWIP as a percentage of invested capital

ATCO Electric – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

2,281,200 2,130,400 1,948,600

4.96 5.08 5.21

5.29 5.31 5.21

3.48 4.62 7.04

FortisAlberta – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

2,905,900 2,695,000 2,499,400

4.81 4.99 5.22

6.77 6.43 6.77

2.21 2.76 2.52

ENMAX – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

1,177,600 1,093,100

995,900

3.93 4.03 4.24

5.17 5.12 5.06

1.96 2.98 5.09

EPCOR – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

987,200 851,000 738,300

5.13 5.00 5.30

4.35 4.30 4.34

3.79 3.57 2.78

ATCO Gas – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

2,313,500 2,144,400 1,997,700

5.36 5.60 5.90

6.35 6.42 6.39

2.45 2.20 2.12

AltaGas – distribution 2016 Rule 005 2015 Rule 005 2014 Rule 005

280,500 244,500 215,800

4.54 4.71 4.90

4.90 4.90 5.12

2.39 2.69 1.48

AltaLink – transmission 2016 Rule 005 2015 Rule 005 2014 Rule 005

6,943,100 5,257,400 5,110,500

4.00 4.11 4.10

4.58 4.50 3.37

4.71 3.49 -1.20

ATCO Electric – transmission 2016 Rule 005 2015 Rule 005 2014 Rule 005

5,235,700 5,197,900 4,630,200

4.77 4.72 4.84

3.59 2.67 2.83

1.44 1.40 1.54

ENMAX – transmission 2016 Rule 005 2015 Rule 005 2014 Rule 005

424,500 392,200 323,500

3.93 4.03 4.24

3.88 3.86 3.72

6.25 5.82

13.13

EPCOR – transmission 2016 Rule 005 2015 Rule 005 2014 Rule 005

671,900 657,700 624,300

5.22 4.93 4.88

3.54 3.40 3.32

2.52 2.50 3.18

ATCO Pipelines – transmission 2016 Rule 005 2015 Rule 005 2014 Rule 005

1,252,700 1,083,300

956,600

5.10 5.29 5.50

5.35 5.14 5.34

12.15 11.04 8.62

Simple average 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.70 4.77 4.94

4.89 4.73 4.68

3.94 3.92 4.21

764. In Table 10 below, the Commission presents additional calculations based on the

information presented in Table 9. There is no simple average or weighted average for gas



transmission utilities presented separately in Table 10 because there is only one gas transmission

utility, i.e., ATCO Pipelines.

Table 10. Additional analysis of information included in Table 9

Utility

Debt cost

%

Depreciation as a percentage of

invested capital


Simple average – overall 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.70 4.77 4.94

4.89 4.73 4.68

3.94 3.92 4.21

Weighted average - overall 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.88 4.58 4.39

3.54 3.24 2.35

Simple average – distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.79 4.90 5.13

5.47 5.41 5.49

2.71 3.14 3.50

Weighted average – distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

5.85 5.77 5.86

2.69 3.16 3.77

Simple average – transmission utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.60 4.62 4.71

4.19 3.91 3.71

5.41 4.85 5.06

Weighted average – transmission utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.22 3.72 3.32

4.12 3.30 1.33

Simple average – electric distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.71 4.78 4.99

5.39 5.29 5.35

2.86 3.48 4.39

Weighted average – electric distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

5.73 5.60 5.72

2.78 3.48 4.39

Simple average – gas distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.95 5.15 5.40

5.62 5.66 5.76

2.42 2.44 1.80

Weighted average – gas distribution utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

6.19 6.26 6.27

2.45 2.25 2.06

Simple average – electric transmission utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.48 4.45 4.51

3.90 3.61 3.31

3.73 3.30 4.16

Weighted average – electric transmission utilities 2016 Rule 005 2015 Rule 005 2014 Rule 005

4.12 3.59 3.14

3.36 2.57 0.68



765. In its credit metric calculations, the Commission adopted the following five parameters:

ROE value, embedded average debt rate, income tax rate, depreciation as a percentage of

invested capital and mid-year CWIP as a percentage of invested capital.

ROE value

766. The Commission has applied an ROE value of 8.5 per cent in its credit metric

calculations, consistent with its findings in Section 8.8.

Embedded average debt rate

767. The simple average of the embedded average debt rates is 4.9 per cent based on the 2014

Rule 005 reports, 4.8 per cent based on the 2015 Rule 005 reports, and 4.7 per cent based on the

2016 Rule 005 reports. These figures demonstrate that the embedded average debt rate is

declining, which is to be expected as the affected utilities continue to retire debt with higher

interest rates and replace it with lower cost debt.

768. The Commission finds that the use of 4.7 per cent for the embedded average debt rate is

reasonable. This figure is between the simple average debt rate for the distribution utilities and

the transmission utilities based on the 2016 Rule 005 reports. Given that the affected utilities are

expected to continue to retire higher interest debt and replace it with lower interest debt, the

Commission considers the use of 4.7 per cent to be conservative.

Income tax rate

769. The Commission determined in Section 9.7.1 that it will continue to analyze credit

metrics using an income tax rate of 27 per cent. The Commission has also determined credit

metrics using an income tax rate of zero, which accounts for the income-tax-exempt utilities, as

well as those utilities that expect to have no taxable income.

Depreciation as a percentage of invested capital

770. The amount of depreciation collected through rates is included in the calculation of the

FFO component of the FFO/debt and FFO coverage ratios.

771. The weighted average depreciation rate as a percentage of invested capital for the

distribution utilities based on the 2016 Rule 005 reports is 5.85 per cent, as shown in Table 10.

The Commission will use this figure in its credit metric calculations for the distribution utilities.

This figure is between the weighted average depreciation rates based on the 2016 Rule 005

reports for the electric distribution utilities (with a figure of 5.73 per cent) and the gas

distribution utilities (with a figure of 6.19 per cent).

772. The weighted average depreciation rate as a percentage of invested capital for the

transmission utilities based on the 2016 Rule 005 reports is 4.22 per cent, as shown in Table 10.

For simplicity and to be conservative, the Commission will round this to 4.2 per cent. This figure

is between the weighted average depreciation rate based on the 2016 Rule 005 reports for the

electric transmission utilities (with a figure of 4.12 per cent) and the rate for ATCO Pipelines of

5.35 per cent.




773. The overall simple average for the utilities based on the 2014, 2015 and 2016 Rule 005

reports does not provide a clear indication of the trend for this parameter. The percentage

decreased from 4.21 using the 2014 Rule 005 report, to 3.92 per cent using the 2015 Rule 005

report. It increased to 3.94 per cent using the 2016 Rule 005 data. The Commission finds the best

way to determine this parameter is to use the simple average of the weighted average values for

2014, 2015 and 2016. For the distribution utilities, the result is 3.21 per cent. For the

transmission utilities, the result is 2.92 per cent. However, the weighted average percentages for

the transmission utilities have increased from 1.33 in 2014, to 3.30 in 2015, to 4.12 per cent in

2016. The Commission finds that in order to reflect this trend, and be conservative, it will

continue to use the figure of five per cent that it used in the 2016 GCOC decision for the

transmission utilities.

774. Based on the credit metric parameters discussed above, the Commission has updated its

credit metric calculations at various equity ratios from the calculations set out in the 2016 GCOC

decision. As previously mentioned, to address the impact of zero income tax on credit metrics,

the Commission has also provided credit metric calculations at various equity ratios, which

reflect an income tax rate of zero. The revised calculations are set out in Table 11, Table 12,

Table 13 and Table 14.

Table 11. Credit metrics compared to equity ratios – Commission calculations – distribution utilities – income tax rate of 27 per cent

EBIT coverage FFO coverage FFO/debt %

Equity ratio (%)

2016 GCOC decision 2018



30 1.9 2.0 3.3 3.4 11.3 11.6

31 2.0 2.0 3.4 3.5 11.6 11.9

32 2.0 2.1 3.4 3.6 11.9 12.2

33 2.1 2.2 3.5 3.6 12.2 12.5

34 2.1 2.2 3.6 3.7 12.5 12.8

35 2.2 2.3 3.6 3.8 12.6 13.2

36 2.2 2.3 3.7 3.8 13.2 13.5

37 2.3 2.4 3.8 3.9 13.5 13.8

38 2.4 2.4 3.8 4.0 13.8 14.2

39 2.4 2.5 3.9 4.1 14.2 14.6

40 2.5 2.6 4.0 4.1 14.6 14.9

41 2.5 2.6 4.1 4.2 14.9 15.3

42 2.6 2.7 4.2 4.3 15.3 15.7

43 2.7 2.8 4.2 4.4 15.8 16.2

44 2.8 2.9 4.3 4.5 16.2 16.6

45 2.8 2.9 4.4 4.6 16.6 17.0



Table 12. Credit metrics compared to equity ratios – Commission calculations – distribution utilities – income tax rate of zero


Equity ratio (%)

2016 GCOC decision,

non-taxable 2018

non-taxable

2016 GCOC decision,

non-taxable 2018

non-taxable

2016 GCOC decision,

non-taxable 2018

non-taxable

30 1.7 1.7 3.3 3.4 11.3 11.6

31 1.7 1.8 3.4 3.5 11.6 11.9

32 1.7 1.8 3.4 3.6 11.9 12.2

33 1.8 1.8 3.5 3.6 12.2 12.5

34 1.8 1.9 3.6 3.8 12.5 12.8

35 1.9 1.9 3.6 3.6 12.8 13.2

36 1.9 2.0 3.7 3.8 13.2 13.5

37 1.9 2.0 3.8 3.9 13.5 13.8

38 2.0 2.0 3.8 4.0 13.8 14.2

39 2.0 2.1 3.9 4.1 14.2 14.6

40 2.1 2.1 4.0 4.1 14.6 14.9

41 2.1 2.2 4.1 4.2 14.9 15.3

42 2.2 2.2 4.2 4.3 15.3 15.7

43 2.2 2.3 4.2 4.4 15.8 16.2

44 2.3 2.3 4.3 4.5 16.2 16.6

45 2.3 2.4 4.4 4.6 16.6 17.0

Table 13. Credit metrics compared to equity ratios – Commission calculations – transmission utilities – income tax rate of 27 per cent


Equity ratio (%)




30 1.9 2.0 2.8 2.9 9.0 9.2

31 2.0 2.0 2.9 3.0 9.2 9.4

32 2.0 2.1 2.9 3.0 9.5 9.7

33 2.1 2.1 3.0 3.1 9.7 10.0

34 2.1 2.2 3.0 3.1 10.0 10.2

35 2.2 2.2 3.1 3.2 10.3 10.5

36 2.2 2.3 3.1 3.3 10.5 10.8

37 2.3 2.3 3.2 3.3 10.8 11.1

38 2.3 2.4 3.3 3.4 11.1 11.4

39 2.4 2.5 3.3 3.4 11.5 11.7

40 2.5 2.5 3.4 3.5 11.8 12.1

41 2.5 2.6 3.5 3.6 12.1 12.4

42 2.6 2.7 3.5 3.7 12.5 12.8

43 2.7 2.7 3.6 3.7 12.8 13.1

44 2.7 2.8 3.7 3.8 13.2 13.5

45 2.8 2.9 3.8 3.9 13.6 13.9



Table 14. Credit metrics compared to equity ratios – Commission calculations – transmission utilities – income tax rate of zero


Equity ratio (%)

2016 GCOC decision,

non-taxable 2018

non-taxable

2016 GCOC decision,

non-taxable 2018

non-taxable

2016 GCOC decision,

non-taxable 2018

non-taxable

30 1.7 1.7 2.8 2.9 9.0 9.2

31 1.7 1.7 2.9 3.0 9.2 9.4

32 1.7 1.8 2.9 3.0 9.5 9.7

33 1.8 1.8 3.0 3.1 9.7 10.0

34 1.8 1.8 3.0 3.1 10.0 10.2

35 1.8 1.9 3.1 3.2 10.3 10.5

36 1.9 1.9 3.1 3.3 10.5 10.8

37 1.9 2.0 3.2 3.3 10.8 11.1

38 2.0 2.0 3.3 3.4 11.1 11.4

39 2.0 2.1 3.3 3.4 11.5 11.7

40 2.1 2.1 3.4 3.5 11.8 12.1

41 2.1 2.1 3.5 3.6 12.1 12.4

42 2.1 2.2 3.5 3.7 12.5 12.8

43 2.2 2.3 3.6 3.7 12.8 13.1

44 2.2 2.3 3.7 3.8 13.2 13.5

45 2.3 2.4 3.8 3.9 13.6 13.9

775. The Commission has undertaken the above calculations in light of the credit metric

findings in Section 9.7.1. The Commission observes that the credit rating metrics required for an

Alberta utility to achieve a credit rating in the A-range have not changed since the 2016 GCOC

decision. Table 15 sets out the guidelines established by the Commission in this section to

achieve a credit rating in the A-range, which assumes a credit rating assessment of “strong” for

the Alberta regulatory environment. The guidelines do not take into account potential

adjustments to the deemed equity ratios that may be necessary in the Commission’s judgment to

take account of the current trend of “negative” noted by credit rating agencies and in particular

by S&P.

776. Table 15 sets out the minimum equity ratio that would be required, in conjunction with an

approved ROE of 8.5 per cent, for distribution and transmission utilities in Alberta with an

income tax rate of 27 per cent, as well as distribution and transmission utilities in Alberta with an

income tax rate of zero per cent, to meet the corresponding credit ratio threshold or range used

by the Commission to establish a credit rating in the A-range. For example, as shown in Table

15, a distribution utility in the 2018 GCOC proceeding that has an income tax rate of 27 per cent,

would require a deemed equity ratio of 30 per cent to achieve an EBIT coverage ratio of 2.0.

That same utility would require a deemed equity ratio somewhere below 30 per cent, in order to

achieve an FFO coverage ratio of 2.0, and an FFO coverage ratio of 3.0. Finally, that same utility

would require a deemed equity ratio below 30 per cent, in order to achieve an FFO/debt ratio of

9.0, while it would require a deemed equity ratio of 35 per cent to achieve an FFO/debt ratio

of 13.0.



Table 15. Commission guidelines for equity ratios to achieve a credit rating in the A-range

Credit metric guideline 2.0 EBIT coverage

2.0 FFO coverage

3.0 FFO coverage

9.0 FFO/debt ratio

13.0 FFO/debt ratio

(%)

2016 distribution utilities – 27 per cent income tax rate 31 Below 30 Below 30 Below 30 36

2018 distribution utilities – 27 per cent income tax rate 30 Below 30 Below 30 Below 30 35

2016 distribution utilities – zero per cent income tax rate 38 Below 30 Below 30 Below 30 36

2018 distribution utilities – zero per cent income tax rate 36 Below 30 Below 30 Below 30 35

2016 transmission utilities – 27 per cent income tax rate 31 Below 30 33 30 44

2018 transmission utilities – 27 per cent income tax rate 30 Below 30 31 30 43

2016 transmission utilities – zero per cent income tax rate 38 Below 30 33 30 44

2018 transmission utilities – zero per cent income tax rate 37 Below 30 31 30 43

777. Based on the results of its credit metric calculations, the Commission continues to find, as

it did in the 2016 GCOC decision, “that absent differences in business risk, the continued

perpetuation of the historical gap in equity ratios between the higher equity ratio awarded to

distribution utilities and the lower equity ratio awarded to transmission utilities is no longer

warranted.”945

9.8 Business risk utility sector analysis

778. In the 2016 GCOC decision, the Commission expressed the following view with respect

to how it accounted for any differences between the Alberta transmission and distribution

utilities as part of its determination of the deemed equity ratios:

… the Commission notes that its credit metric calculations do not support the

continuation of a 400 bps difference in the awarded deemed equity ratios based on

financial risk. It is also unclear that a difference of any amount remains warranted using

only a credit metric financial risk analysis. From a business risk perspective, the

Commission agrees that there are differences in rate regulation (for example: PBR versus

cost-of-service rate regulation) and depreciation rate differences between transmission

and distribution utilities, and other business risk differences, such as the method of

recovery of fixed costs, although this is somewhat mitigated for the gas distribution

utilities under PBR which accounts for actual changes in customer usage. Accordingly,

the Commission will balance the financial risks as examined in the credit metric

calculations and business risks including utility sector business risks, in arriving at its

final deemed equity ratio determinations.946

945 Decision 20622-D01-2016, paragraph 433. 946 Decision 20622-D01-2016, paragraph 533.



779. In this proceeding, the Commission asked each of Dr. Carpenter, Dr. Villadsen,

Mr. Buttke, Mr. Coyne, Mr. Hevert, Mr. Johnson, Mr. Madsen, Mr. Bell and Dr. Cleary to

provide their respective views on the relative riskiness of the Alberta transmission and

distribution utilities.

780. Dr. Carpenter commented that there is insufficient data or granularity in order to make

distinctions between the relative riskiness, and did not object to the Commission’s decision in

2016 to grant deemed equity ratios that were essentially equivalent for all the Alberta utilities,

with the exception of ENMAX and AltaGas.947 Dr. Villadsen concurred with Dr. Carpenter, and

Mr. Buttke did not offer an opinion.948

781. Dr. Villadsen recommended that the relative deemed equity ratios from the 2016 GCOC

decision stay in place, because nothing in her analysis suggests that the criteria used to determine

the relative deemed equity ratios of the distribution and transmission utilities has changed since

the 2016 GCOC decision.949

782. Mr. Coyne expressed the view that Alberta transmission and distribution utilities are

essentially the same, and he did not distinguish between them on a risk basis when determining

his recommended deemed equity ratio.950

783. Mr. Hevert stated that while he would not necessarily disagree with the Commission’s

previous recognition of somewhat more risk associated with the distribution utilities, he did

recommend the same deemed equity ratio for the Alberta transmission and distribution utilities

as part of this proceeding.951

784. Mr. Johnson indicated that, on average, the Alberta transmission and distribution utilities

have essentially the same risk. He stated that within the Alberta distribution utilities, ATCO Gas

is less risky.952 Mr. Madsen considered that the risk profile of the distribution utilities has

decreased relative to the risk profile of the transmission utilities in the last number of years.953

785. Mr. Bell stated he does not see the Alberta transmission and distribution utilities as

materially different. Dr. Cleary indicated that based on his quantitative analysis of the CV

(EBIT/sales), the differences in business risk between the Alberta transmission and distribution

utilities are not as pronounced as argued in previous proceedings.954

Commission findings

786. In this decision, the Commission will balance the financial risks as examined in the credit

metric calculations, and its analysis of business risks, including utility sector business risks, in

arriving at its final deemed equity ratio determinations. The Commission notes that no parties

947 Transcript, Volume 4, pages 696-697. 948 Transcript, Volume 4, pages 697-698. 949 Exhibit 22570-X0193.01, A92. 950 Transcript, Volume 5, pages 980-981. 951 Transcript, Volume 6, pages 1251-1252. 952 Transcript, Volume 7, pages 1347-1348. 953 Exhibit 22570-X0701.01, CCA-AUC-2018JAN26-014. 954 Transcript, Volume 10, pages 2157-2159.



identified any significant business risks differences between the distribution utilities and the

transmission utilities that would justify different deemed equity ratios for the two sectors.

9.9 Equity ratio adjustments for income-tax-exempt or non-taxable utilities

787. Mr. Coyne stated that it is necessary to add at least 200 bps to the deemed equity ratio of

a non-taxable utility to achieve the same level of equity return that is approved for taxable

utilities. He submitted this will provide the non-taxable utilities compensation for the additional

risk they bear relative to the taxable utilities. Mr. Coyne submitted that denying this 200 bps

adder effectively awards a higher return to taxable utilities for what is otherwise the same level

of risk. He stated that this violates the comparability principle of the fair return standard.955

788. Mr. Coyne recommended a 42 per cent deemed equity ratio for non-taxable utilities.956

He submitted that the non-taxable utilities require higher deemed equity ratios to achieve the

same credit metrics as the taxable utilities.957

789. Mr. Coyne indicated that a taxable utility holds a cash-flow advantage over a non-taxable

utility, because it receives recovery of income tax expense in its revenue requirement. He

submitted that the regulatory benefits arising from income tax recovery could be equalized by

adjusting the deemed equity ratios such that taxable and non-taxable utilities achieve the same

credit metrics.958

790. Mr. Coyne submitted that while taxable utilities in Alberta bear very little risk for their

exposure to income taxes, they realize significant financial benefits. He referred to Mr. Bell’s

submission that non-taxable utilities do not have the income tax shield that is available to taxable

utilities to cushion the after-tax impact of changes in deductible costs. Mr. Coyne explained that

if the O&M expenses of a taxable utility increase above the approved amount then, all else equal,

the income taxes decrease. This is not the case for non-taxable utilities. Mr. Coyne submitted this

discrepancy should be reflected in a higher equity ratio for non-taxable utilities.959

791. Mr. Madsen argued that the 200 bps increase requested by ENMAX as an income-tax-

exempt utility is not required to support credit metrics for 2018 to 2020. He stated this is no

different than other forms of credit metric relief, and it should only be provided if it is necessary

for the utility to maintain a credit rating in the A-range, after all other non-equity based credit

metric relief measures have been implemented.960 The CCA agreed with Mr. Madsen’s

submissions.961

Commission findings

792. The Commission addressed this issue in the 2016 GCOC decision, and determined that a

200 bps adder to the deemed equity ratios for income-tax-exempt or non-taxable utilities was not

warranted. In this proceeding, Mr. Coyne did not distinguish between income-tax-exempt

utilities, such as ENMAX and EPCOR, and non-taxable utilities, which could apply to utilities

955 Exhibit 22570-X0131, PDF pages 105-106. 956 Exhibit 22570-X0131, PDF page 106. 957 Exhibit 22570-X0131, PDF pages 101-102. 958 Exhibit 22570-X0131, PDF pages 105-106. 959 Exhibit 22570-X0775, PDF pages 61-62. 960 Exhibit 22570-X0557, paragraphs 114-115. 961 Exhibit 22570-X0888, paragraph 470.



that are required to pay income taxes, but currently have no taxable income. The Commission

considers that because Mr. Coyne presented evidence on behalf of ENMAX, his

recommendation for a 200 bps adder applies to the income-tax-exempt utilities.

793. Mr. Coyne submitted that there is a difference in credit metrics between the income-tax-

exempt utilities and the taxable utilities, and the addition of a 200 bps adder to the deemed equity

ratio of the income-tax-exempt utilities would alleviate this problem. The Commission notes that

the FFO/debt and the FFO coverage ratios are not affected by income tax status, whereas the

EBIT coverage, EBITDA coverage and debt/EBITDA ratios are. In the 2016 GCOC decision,

the Commission agreed with parties that the most important credit ratio to focus on was the

FFO/debt ratio.962 No evidence was submitted in this proceeding that would alter the

Commission’s view on this matter. Therefore, the Commission remains of the view that the most

important credit metric to focus on is the FFO/debt ratio.

794. As part of its credit metric analysis in Section 9.7.2, the Commission looked at the credit

metrics that would be achieved by income-tax-exempt distribution utilities and income-tax-

exempt transmission utilities. The Commission agrees that there are differences between the

taxable utilities and the income-tax-exempt utilities with respect to the EBIT/coverage and

EBITDA coverage ratios. However, the Commission notes that at a deemed equity ratio of

37 per cent, the EBIT coverage and EBITDA coverage ratios for income-tax-exempt distribution

utilities of 2.0 and 3.9, respectively, are within the applicable A-range thresholds for DBRS and

S&P. The Commission also notes that at a deemed equity ratio of 37 per cent, the EBIT coverage

and EBITDA coverage ratios for income-tax-exempt transmission utilities of 2.0 and 3.3,

respectively, are within the A-range thresholds for DBRS and S&P.

795. Mr. Coyne submitted that denying the 200 bps adder awards a relatively higher return to

taxable utilities for what is otherwise the same level of risk. He based this on his judgment that

the taxable utilities in Alberta bear very little risk for their exposure to income taxes. The

Commission disagrees with Mr. Coyne that the taxable utilities in Alberta bear very little risk for

their exposure to income taxes.

796. As described in Section 5.4, every one of the taxable utilities has requested that one or

more deferral accounts be established for a variety of aspects regarding income tax. These range

from changes in statutory income tax rates and capital cost allowance rates, to protection against

income tax reassessments, to material amendments to income tax legislation. The Commission

considers that if the taxable utilities perceived very little risk relating to income taxes, they

would not be requesting these deferral accounts. The Commission finds that there are business

risks related to income tax that are faced by the taxable utilities, and not faced by the income-tax-

exempt utilities. This difference in business risk must be considered when assessing the

disadvantage the income-tax-exempt utilities have in the area of certain credit metrics.

797. Mr. Coyne and Mr. Bell discussed the income tax shield that is available to taxable

utilities. Mr. Coyne explained that if operating expenses for a taxable utility increase above the

approved amount then, all else equal, the income taxes decrease. The Commission agrees with

this statement, but it notes that the reverse situation applies as well. If the operating expenses of

a taxable utility decrease below the approved amount then, all else equal, the income taxes

962 Decision 20622-D01-2016, paragraph 563.



increase. Any operating expense savings for an income-tax-exempt utility flow through 100 per

cent to net income.

798. Based on the foregoing, the Commission finds that a 200 bps adder to the deemed equity

ratio for the income-tax-exempt utilities is not warranted.

9.10 Equity ratio adjustments for ENMAX

799. In this section, the Commission will review the business risk evidence for ENMAX.

800. Mr. Coyne noted that the deemed equity ratio for ENMAX Distribution in 2015 was

40 per cent, and dropped to 36 per cent for 2017. He submitted that in the 2016 GCOC decision,

the Commission cited no substantive change in the risk of ENMAX Distribution that would

warrant such a dramatic reduction.963

801. Mr. Coyne reviewed the business risk profile of ENMAX. Based on his review, he stated

there are material differences between ENMAX and the average Alberta utility, but his belief is

that these differences do not warrant an explicit adjustment from the approved ROE and deemed

equity ratio he recommended for the income-tax-exempt utilities in Alberta. He submitted there

is nothing to suggest that ENMAX should be considered to be lower risk than the other utilities

in Alberta, and indeed ENMAX has higher risk in many respects.964

Commission findings

802. The Commission set out its reasons for establishing the current deemed equity ratio for

ENMAX in the 2016 GCOC decision, and in Decision 22211-D01-2017. The main reason for the

300 bps reduction in ENMAX Distribution’s deemed equity ratio approved on an interim basis in

the 2016 GCOC decision was the elimination of the 200 bps adder that had been previously

awarded because of ENMAX’s income-tax-exempt status. The reason for the 100 bps reduction

in Decision 22211-D01-2017 was the deviation between ENMAX’s actual equity ratios and its

deemed equity ratios, and its apparent ability to operate at a significantly lower year-end equity

ratio without impairment to its ongoing operations, its financial integrity or its ability to raise

capital.965

803. The Commission acknowledges the submission of Mr. Coyne that ENMAX has

committed to maintaining an actual equity ratio that is consistent with its deemed equity ratio. In

reviewing the 2016 Rule 005 reports for ENMAX Transmission and ENMAX Distribution, the

Commission notes that the actual year-end ratio for both was 37 per cent.966 The Commission

agrees with Mr. Coyne that there is nothing to suggest that ENMAX should be considered lower

risk than the other utilities in Alberta. Therefore, the Commission finds that the deemed equity

ratio for ENMAX should be the same as the other affected utilities, with the exception of

AltaGas.

963 Exhibit 22570-X0131, PDF page 84. 964 Exhibit 22570-X0131, PDF page 109. 965 Decision 22211-D01-2017, paragraphs 79-80. 966 Exhibit 22570-X0139, worksheet 2.2 TT. Exhibit 22570-X0138, worksheet 2.2 DT.



9.11 Determination of Commission-approved deemed equity ratios

804. In this section on capital structure, the Commission started with a review of generic

business risks. This included an overall assessment of the business risks of the affected utilities,

in Section 9.3.1. The Commission considered that the favourable financial performance of the

affected utilities over the 2005 to 2016 period is support for assessing the utilities in Alberta as

having low financial risk.

805. The generic business risk review continued in Section 9.3.2. There, the Commission

looked at the specific issues identified by the utilities that, in their submission, have increased

their regulatory risk since the 2016 GCOC decision. These issues were (1) the 2018 to 2022 PBR

term; (2) the Commission’s UAD decision and the related issue of asset utilization; (3) the

increase in customer contributions; (4) regulatory lag; and (5) clean energy initiatives. The

Commission found no increase in business risk since the 2016 GCOC decision as a consequence

of any of these five specific issues.

806. In Section 9.3.3, the Commission reviewed the submissions on the business risk

comparisons between the affected utilities and utilities in other jurisdictions, primarily in the

U.S. The Commission found no persuasive evidentiary support for the conclusion that regulatory

risk in the U.S. is less than that for utilities in Alberta.

807. The Commission examined evidence comparing the deemed equity ratios it awards to the

deemed equity ratios awarded in other jurisdictions in Section 9.3.4 and found there are reasons

why the deemed equity ratios awarded in Alberta cannot be compared to those awarded by U.S.

regulators. The Commission also found that the deemed equity ratio it awarded in the 2016

GCOC decision is comparable to those in other Canadian jurisdictions.

808. In Section 9.4, the Commission considered the submissions from Mr. Hevert regarding

industry financing practices, and then reviewed the submissions of FortisAlberta regarding

equity attraction. The Commission found that the utilities will always face some refinancing

risks, and assessing long-term refinancing risk for the utilities in Alberta will involve long-term

assumptions about bond markets, which could change substantially over the ensuing years.

809. The Commission further addressed the considerations brought forward by FortisAlberta

on equity attraction in Section 9.5. Among other findings, the Commission considered that

FortisAlberta’s equity investor would likely assess not only the approved ROE but also the actual

ROEs that have been achieved by FortisAlberta, which averaged 10.2 per cent over the years

2010 to 2016.

810. The Commission’s analysis of financial risk, focusing on the credit metrics required to

achieve an A-range credit rating, is set out in Section 9.7.

811. Based on the information in Table 11 and Table 13, the Commission notes that an

average distribution utility and an average transmission utility, with an income tax rate of 27 per

cent, would meet all the credit metric guidelines of the Commission, with an ROE of 8.5 per

cent, at a deemed equity ratio of 30 per cent. An average distribution and transmission utility that

is either income-tax-exempt or currently non-taxable, would meet the FFO/debt and FFO

coverage credit metric guidelines of the Commission, with an ROE of 8.5 per cent, at a deemed

equity ratio of 30 per cent. The Commission’s EBIT coverage credit metric guideline would be

met with a 36 per cent deemed equity ratio at an ROE of 8.5 per cent for those distribution



utilities that have no income tax expense, and with a 37 deemed equity ratio for the transmission

utilities that have no income tax expense.

812. In Section 9.8, the Commission examined whether there are business risk differences

between the distribution utilities and the transmission utilities that would warrant different

deemed equity ratios. The Commission concluded that no such finding is warranted.

813. In Section 9.9, the Commission reviewed the recommendation of Mr. Coyne that the

income-tax-exempt utilities should receive a 200 bps adder to their deemed equity ratio. . Based

on its findings in that section, the Commission determined that no adder was warranted.

814. Finally, in Section 9.10, the Commission examined the submissions regarding the

deemed equity ratio for ENMAX. Based on its review of the evidence, the Commission found

that the deemed equity ratio for ENMAX should be the same as for the other utilities in Alberta,

with the exception of AltaGas.

815. Considering all the information and findings set out in this capital structure section, the

Commission finds that no change is required to the deemed equity ratio set out in the 2016

GCOC decision, with the exception of the deemed equity ratio for ENMAX, as discussed in

Section 9.10, and the deemed equity ratio of AltaGas, which will be addressed in the following

section. The Commission has determined that a deemed equity ratio of 37 per cent for both

distribution and transmission utilities, with the exception of AltaGas, including those which pay

income tax and those which currently are income tax exempt or do not currently pay income tax,

satisfies the fair return standard when combined with an 8.5 per cent approved ROE for 2018 to

2020, and will enable the affected utilities to maintain a credit rating in the A-range.

10 Determination of Commission-approved deemed equity ratio for AltaGas Utilities

Inc.

816. In this section, the Commission will determine the deemed equity ratio for AltaGas,

considering the determinations previously made in this decision, as well as the evidence

regarding the business risk of AltaGas and the submissions regarding the actual credit rating of

AltaGas.

Business risk

817. AltaGas submitted that its business risk is higher than the benchmark Alberta utility,

which has been recognized in previous GCOC decisions. It indicated that the previously awarded

400 bps adder continues to be appropriate to reflect its relative risk to the benchmark Alberta

utility. AltaGas stated that this is supported by (1) its size and geographically dispersed system;

(2) its gas supply risk; (3) the support of all parties in the proceeding; (4) previous Commission

decisions; and (5) the application of the stand-alone principle.967

818. Dr. Carpenter submitted that the 400 bps adder to the deemed equity ratio for AltaGas be

maintained.968 He indicated that AltaGas faces unusual supply risk because of the risk that some

of the older lateral supply pipelines it uses but does not own may be shut in. This will require

967 Exhibit 22570-X0898, paragraphs 4-5. 968 Exhibit 22570-X0131, A6.



AltaGas to either obtain alternative supplies, or take on financial responsibility for owning and/or

maintaining these old lateral supply pipelines. He added that if the utility has to take on

responsibility for the aging supply infrastructure, this could also elevate the operating risk for

AltaGas. Dr. Carpenter stated that these supply risks are significant for AltaGas, because of its

small size and dispersed service territory.969

Actual credit rating

819. In a letter dated January 24, 2018, the Commission invited parties to comment on an

issue that had arisen in Proceeding 23010: AltaGas Utility Group Inc. – Application for the Sale

and Transfer of Capital Stock. In that letter, the Commission articulated the issue as follows:

“Should a utility incapable of raising debt at an A rating receive a deemed equity ratio and

deemed return on equity premised upon an A credit rating?”970

820. The Commission issued Decision 23010-D01-2018971 on January 30, 2018, in which it

noted the following:

33. The Commission in an IR raised the issue of an apparent disconnect between the

equity thickness the Commission awarded AltaGas Utilities based on a credit rating of

A category in the GCOC decisions and the cost of debt that is flowed to its customers

based on an investment grade credit rating of BBB of AltaGas Ltd. AltaGas Group, in its

response stated:

AUI’s current rates reflect the interest rates on the debt that was present during the

2012 test year as adjusted in compliance filing, and further adjusted for the effects of

the PBR formula and capital tracker proceedings during the years 2013-2017.

…

It is important to note, that the 2016 GCOC Decision did not direct AUI, or any

Alberta utility, to modify its interest expense from those approved within the first

generation PBR Decision 2012-237. AUI customers are paying rates that are fully

reflective of past Commission decisions – including equity ratio and ROE as

determined by the Commission in its GCOC proceedings and debt rates tested

separately by the Commission for prudence.

34. AltaGas expressed its view that any relationship between a utility’s actual credit

rating and the resulting cost of debt on one hand, and the findings in the Commission’s

GCOC decisions regarding the allowed ROE and equity thickness awarded to the utility

based on an A category credit rating on the other hand, requires a wider forum. The

Commission agrees and considers the 2018 GCOC proceeding to be the correct forum to

address this issue. [footnotes omitted]972

821. AltaGas submitted that the Commission has already addressed the issue articulated in the

January 24, 2018 letter, through the robust ratemaking principles and processes that have been

established over the last decade or more. It commented that the Commission has adopted

969 Exhibit 22570-X0186, A14-A15. 970 Exhibit 22570-X0616. 971 Decision 23010-D01-2018: AltaGas Utility Group Inc., Application for the Sale and Transfer of Capital Stock,

Proceeding 23020, January 30, 2018. 972 Decision 23010-D01-2018, paragraphs 33-34.



principles to ensure that customers of a particular utility only pay debt costs that are deemed as

prudent. AltaGas pointed out that in its decision on AltaGas’s 2010-2012 general rate application

(GRA),973 the Commission reduced AltaGas’s allowed cost of debt rate on two debentures, based

on a prudency review.974

822. AltaGas submitted that the Commission has, and can continue to, address any issues

associated with the cost of debt, including the issue identified by the Commission in its

January 24, 2018 letter, through a prudency review in a general rate case, rather than through

changes to capital structure.975 For the PBR utilities, AltaGas indicated this prudence review

takes place in a rebasing proceeding for going-in rates.976 AltaGas noted that it’s going-in debt

costs for the 2018-2022 PBR term reflect an average embedded rate of 4.46 per cent. It indicated

that it will bear the incremental cost for any debt it issues during the 2018-2022 PBR term at

rates in excess of 4.46 per cent. AltaGas stated that in the next PBR rebasing proceeding, the

Commission will determine the prudency of any debt issuances made by AltaGas during the

2018-2022 PBR term.977

823. Following its consideration of all the submissions made by parties in response to its

January 24, 2018 letter, the Commission advised the parties on February 9, 2018 as follows:

6. The Commission agrees with the submissions of Fortis, AltaLink, EPCOR and the

ATCO Utilities that the specific issue raised in the Commission’s letter of January 24,

2018, currently relates only to AltaGas and does not need to be considered in the 2018

GCOC proceeding for other companies. However, the Commission will address the issue

as it relates to AltaGas’s deemed equity ratio and return on equity to be approved in this

proceeding, and considers that this is within the existing scope of the proceeding.

7. Further, the Commission notes that the specific issue identified in relation to AltaGas

may also be considered, in part, in the context of the Commission’s practice of

maintaining credit ratings in the A category for the utilities in Alberta. This matter was

explored by the Commission in interrogatories and may be explored further during the

oral hearing.978

824. In argument, AltaGas submitted its customers are not harmed by its current debt

practices, because they pay for debt at rates that have been determined to be prudent by the

Commission.979 AltaGas considered that the evidence of Dr. Carpenter and Dr. Villadsen

regarding the fair return standard is relevant in addressing this issue.980 AltaGas stated that the

Commission has historically acknowledged and considered, in past GCOC decisions, AltaGas’s

access to BBB-rated debt and its unique business risks when deciding a fair return and

973 Decision 2012-091: AltaGas Utilities Inc., 2010-2012 General Rate Application – Phase I, Proceeding 904,

Application 1606694-1, April 9, 2012. 974 Exhibit 22570-X0652, PDF page 1. 975 Exhibit 22570-X0652, PDF page 2. Exhibit 22570-X0783, paragraph 51. 976 Exhibit 22570-X0652, PDF page 2. 977 Exhibit 22570-X0652, PDF page 2. 978 Exhibit 22570-X0658, paragraphs 6-7. 979 Exhibit 22570-X0921, paragraph 11. 980 Exhibit 22570-X0783, paragraph 49.



establishing a deemed equity ratio for AltaGas. It added that its cost of debt has historically been

addressed for prudence in its GRAs and annual capital tracker true up applications.981

825. The CCA submitted that the Commission must determine whether it remains in the public

interest to continue to target an A-range credit rating for AltaGas considering the associated cost

of doing so and that the benefit of reduced debt costs is not received.982

826. Dr. Cleary described the situation that is occurring with AltaGas as not desirable. He

indicated that customers are bearing the additional cost of the deemed equity ratio increase to

maintain the A-range credit rating, and customers are paying the higher interest rates associated

with a sub-A credit rating. He commented that AltaGas is being rewarded for not being able to

maintain the company’s financial health.983

827. Mr. Bell stated that it is incumbent upon the Commission to ensure that AltaGas’s rates

only include the cost of debt that would be attributable to an A-range credit rating, and that there

must be symmetry.984

828. AltaGas commented that any credit metric analysis that targets an A-range credit rating

does not guarantee an A-range credit rating for any particular utility. It noted this is the case in

its situation, especially when its debenture issues are small. AltaGas stated that its embedded cost

of debt is in the middle of the range for the Alberta utilities.985 This observation was shared by

Dr. Villadsen.986

829. AltaGas submitted that the fair return standard, comprised of the deemed equity ratio and

the approved ROE, should not be conflated with the cost of debt. It suggested that while credit

metrics provide a useful baseline for assessing the deemed equity ratio, business risks must also

be assessed to ensure the fair return standard is met. AltaGas stated its business risks have not

declined.987

830. AltaGas stated that if the Commission still views this issue as being outstanding, it

requests an opportunity to address the merits of any outstanding matter in a proper, considered

and procedurally fair manner.988

Commission findings

831. As a preliminary matter, the Commission rejects AltaGas’s submission that the issue

articulated in the Commission’s January 24, 2018 letter has already been addressed.

832. AltaGas stated that in prior GCOC and GRA decisions, the Commission and its

predecessor have historically acknowledged and considered AltaGas’s access to BBB-rated debt

and its unique business risks when deciding a fair return and establishing a deemed equity ratio

981 Exhibit 22570-X0848, PDF pages 1-2. 982 Exhibit 22570-X0888, paragraph 299. 983 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005. 984 Transcript, Volume 10, page 2160. 985 Exhibit 22570-X0898, paragraphs 38-39. 986 Transcript, Volume 3, page 544. 987 Exhibit 22570-X0898, paragraphs 43-44. 988 Exhibit 22570-X0921, paragraph 21.



for AltaGas.989 Further, AltaGas indicated that its cost of debt has historically been addressed for

prudence, separate and apart from ROE and capital structure matters, in AltaGas’s GRAs.

Accordingly, AltaGas submitted that its access to BBB-rated debt and its unique business risks

have already been accounted for in determining a fair return for AltaGas.

833. The level of AltaGas’s debt costs is not new. However, the issue before the Commission

in this proceeding is whether the Commission should continue to establish a deemed equity ratio

for AltaGas that, when combined with the approved ROE, will achieve target credit ratings in the

A-range, given that AltaGas’s customers do not receive the benefit of debt financing obtained at

A-range credit-rating levels. This issue was highlighted in Proceeding 23010, and is one that the

Commission is satisfied has not been specifically examined in any prior proceeding that

addressed AltaGas’s cost of capital.

834. The Commission also rejects AltaGas’s contention that, if the Commission still views this

issue as outstanding, AltaGas should be afforded a further opportunity to address the merits of

the matter in a proper, considered and procedurally fair manner before any changes are made to

AltaGas’s capital structure. The inability of AltaGas to obtain debt at A-range credit-rating levels

was specifically identified as an issue in this proceeding in the Commission’s January 24, 2018

letter, following Proceeding 23010. In that proceeding, AltaGas stated that this issue was best

addressed in a forum other than Proceeding 23010.990

835. In subsequent correspondence issued in this proceeding on February 9, 2018, the

Commission expressly stated that it would “address the issue as it relates to AltaGas’s deemed

equity ratio and return on equity to be approved in this proceeding, and considers that this is

within the existing scope of the proceeding.”991 It added, “Further, the Commission notes that the

specific issue identified in relation to AltaGas may also be considered, in part, in the context of

the Commission’s practice of maintaining credit ratings in the A category for the utilities in

Alberta. This matter was explored by the Commission in interrogatories and may be explored

further during the oral hearing.”992 Finally, the Commission observes that this issue was explored

in the oral hearing993 without objection from AltaGas, and AltaGas addressed this issue in its

rebuttal evidence filed on February 28, 2018.994 The Commission is satisfied that AltaGas had

reasonable notice of the Commission’s intention to address this issue in this GCOC proceeding

as well as an opportunity to make submissions in response, which it did.

836. Turning to the substantive issue, the Commission accepts the evidence presented that the

business risk of AltaGas is greater than that of the other utilities in Alberta. This, on its own,

suggests that the deemed equity ratio for AltaGas should be greater than the deemed equity ratio

for the other utilities, if all were targeted to achieve A-range credit ratings. However, the

Commission has determined that the inability of AltaGas to raise debt at A-range credit-rating

levels, and the uncertainty with respect to AltaGas’s future debt costs, warrants a downward

adjustment to the deemed equity ratio of AltaGas, relative to that approved for the other utilities.

989 Exhibit 22570-X0898, paragraph 35. Exhibit 22570-X0848, PDF page 2. 990 Exhibit 22570-X0616, PDF page 3. 991 Exhibit 22570-X0658, paragraph 6. 992 Exhibit 22570-X0658, paragraph 7. 993 See, for example, Transcript, Volume 3, starting at page 542. 994 Exhibit 22570-X0783, paragraphs 48-50.



837. The Commission agrees with AltaGas that this issue does not relate to the prudence of its

long-term debt rates. Rather, the issue relates to the Commission’s duty to set a fair return for

AltaGas as an element of the just and reasonable rates to be paid by its customers.

838. The Commission has taken specific note of the evidence in this proceeding with respect

to AltaGas’s inability to obtain debt at A-range credit-rating levels. In addition, as discussed in

Decision 23010-D01-2018, there is uncertainty with respect to the cost of debt that AltaGas’s

new parent can access (debt which will be mirrored down to AltaGas).995 In this GCOC

proceeding, AltaGas indicated that it “has always obtained debt financing from its parent

[AltaGas Ltd.], which has never had access to A grade debt. [emphasis added]”996 Moreover,

there is no evidence to suggest that AltaGas will be able to issue new debt at A-range credit-

rating levels.

839. Further, AltaGas acknowledged that on a stand-alone basis, it might not be able to

achieve an A-range credit rating, because of the small size of its debenture issues. Dr. Villadsen

agreed with this. Mr. Buttke indicated that in order to go into the bond index in Canada, the issue

size has to be a minimum of $100 million and, by not being in the bond index, one can lose

access to a lot of buyers.997 The Commission notes that the largest debt issuance by AltaGas since

2009 was $45 million, which would not qualify it for inclusion in the bond index. The

Commission therefore agrees that AltaGas would likely not be able to achieve an A-range credit

rating on a stand-alone basis.

840. Because AltaGas is unable to access lower cost debt that is associated with an A-range

credit rating, coupled with the uncertainty of its future debt costs, the Commission considers that

AltaGas’s deemed equity ratio should be lowered. Otherwise, as Dr. Cleary stated, “consumers

bear the costs of both the additional cost of the increase in equity thickness and the cost of

paying interest rates above those for an A-rated utility.”998

841. The Commission notes that a sizeable reduction to AltaGas’s deemed equity ratio would

be required to target credit ratings in the BBB-range, which would result in AltaGas having a

significantly lower deemed equity ratio than the other affected utilities. The Commission does

not consider that a reduction of this magnitude is reasonable, particularly given its continued

findings regarding AltaGas’s business risk, as compared to the other affected utilities.

Historically, the Commission has awarded a higher deemed equity ratio to AltaGas than the other

affected utilities, to recognize its relatively higher risk. However, the Commission has

determined that some reduction in equity thickness is warranted to allow for greater symmetry

between the credit rating associated with AltaGas’s debt and its equity thickness.

842. In the 2016 GCOC decision, the Commission awarded AltaGas a 41 per cent deemed

equity ratio to recognize its risk, relative to the other affected utilities. In this decision, given the

Commission’s findings that a reduction in the deemed equity ratio is warranted to recognize

995 In Decision 23010-D01-2018, paragraph 30, the Commission acknowledged that in the event that AltaGas’s

new parent, AltaGas Utility Holdings (Pacific) Inc., is unable to obtain the investment grade credit rating, it was

confirmed that for any new debt issued by AltaGas, the interest expense to be recovered from AltaGas will be

based on the interest rate available at the time for investment grade (DBRS, BBB (low) rate debt). 996 Exhibit 22570-X0820, PDF page 5. 997 Transcript, Volume 3, page 545. 998 Exhibit 22570-X0675, UCA-AUC-2018JAN26-005.



AltaGas’s inability to obtain debt at A-range credit-rating levels, the Commission finds that a

deemed equity ratio of 39 per cent for AltaGas for 2018 to 2020 is reasonable. The Commission

considers that the resulting deemed equity ratio balances AltaGas’s higher business risk

compared to the other affected utilities, with a reduction to account for the actual credit rating

associated with AltaGas’s debt.

843. Given the evidence on the record of this proceeding, the Commission has determined that

a deemed equity ratio of 39 per cent for AltaGas, when combined with an 8.5 per cent ROE for

2018, 2019 and 2020, satisfies the fair return standard.

11 Other areas included in scope of the proceeding

11.1 Maintaining actual equity ratio in line with deemed equity ratio

844. In this proceeding, the Commission asked AltaGas,999 AltaLink,1000 the ATCO Utilities,1001

ENMAX,1002 EPCOR1003 and FortisAlberta1004 to provide their opinions about whether

comparisons between actual equity ratios and the approved deemed equity ratio should be done

using actual mid-year data, or actual year-end data, or both. Certain of the utilities1005 submitted

that year-end data should be used; others1006 submitted that mid-year data should be utilized; and

a couple1007 indicated that both year-end and mid-year data should be used.

845. Mr. Thygesen agreed with FortisAlberta’s submission that utilities should consistently

attempt to align their actual equity ratios with the deemed equity ratios approved by the

Commission.1008 He contended that the utilities do this every day, and he suggested this could be

accomplished by using short-term debt until it builds up to a certain level, and then converting

that level of short-term debt to long-term debt.1009

846. Mr. Buttke stated that the Commission should not regulate the cash management and debt

issuance practices of the utilities on a tactical level. He commented that having low liquidity in

a capital intensive business would be a greater concern than Mr. Thygesen’s concern about

having excess liquidity from time to time. Mr. Buttke indicated that maintaining liquidity and

maintaining the ability to fund the utility’s capital needs is the primary role of the treasurer, and

while minimizing the cost is an important part of that role, it is a secondary role.1010

847. AltaGas suggested that short-term debt does not generally comprise a permanent source

of capital, and short-term debt balances comprise a small percentage of financing requirements

when compared to long-term debt. It submitted that placing prescriptive restrictions on utilities,

999 Exhibit 22570-X0512, AUI-AUC-2017NOV21-002. 1000 Exhibit 22570-X0438, AML-AUC-2017NOV21-002. 1001 Exhibit 22570-X0352, ATCOUTILITIES-AUC-2017NOV21-002. 1002 Exhibit 22570-X0286, EPC-AUC-2017NOV21-002. 1003 Exhibit 22570-X0434, EDTI-AUC-2017NOV21-002. 1004 Exhibit 22570-X0462, FAI-AUC-2017NOV21-002. 1005 The ATCO Utilities and ENMAX. 1006 AltaGas and FortisAlberta. 1007 AltaLink and EPCOR. 1008 Exhibit 22570-X0551, paragraph 197. 1009 Exhibit 22570-X0551, paragraphs 201-202. 1010 Exhibit 22570-X0749, A63.



by forcing them to rebalance debt and equity ratios on a monthly basis, has the potential to

increase costs and decrease efficiencies by having to administer what would be relatively small

amounts of long-term debt and equity issues on a monthly basis.1011

848. The ATCO Utilities indicated their long-term debt procurement practice is that financing

requirements are determined during the year using annual capital expenditure requirements,

while targeting the deemed equity ratio at year-end. They stated that their monthly cash balances

can fluctuate widely from month to month for a variety of reasons. The ATCO Utilities

contended that a daily monitoring regime would introduce inflexibility into its treasury

practices.1012

849. ENMAX submitted it is unrealistic to expect a utility’s capital structure to be exactly

aligned with the approved deemed capital structure every day of the year.1013

Commission findings

850. The Commission finds Mr. Thygesen’s submission that the utilities should consistently

attempt to align their actual and deemed equity ratios by using short-term debt until it builds to

a certain level is not realistic. The Commission considers that the continual rebalancing of debt

and equity ratios should not take precedence over a utility’s cash management practice.

851. Mr. Thygesen recommended that the utilities should allow short-term debt to build up to

a certain level and then convert it to long-term debt. He did not provide further detail on this

proposal. The Commission considers that a recommendation of this nature must consider how

the level of short-term debt should be established for every utility, and how this would factor in

the capital expenditure requirements for each year. In addition, this type of recommendation

would need to include details on how often a utility should issue long-term debt during a year,

and an assessment of the costs and benefits associated with varying levels of short-term debt and

long-term debt, including the costs associated with issuing long-term debt in the market.

852. Based on the foregoing, the Commission will not require the affected utilities to

consistently attempt to align their actual equity ratios with their deemed equity ratios, by altering

their cash management practices as recommended by Mr. Thygesen.

11.2 Reporting of information in Rule 005

853. Mr. Thygesen recommended that the affected utilities be required to report monthly cash

and cash equivalent levels, in Section 2 of their Rule 005 reports. He submitted this would be a

cost-effective method to ensure that the actual equity ratios are maintained at the approved

levels.1014 He further recommended that the utilities also report monthly short-term debt balances

in Section 2 of their Rule 005 reports, which he submitted would help to ensure that the utilities

are maintaining actual equity ratios in line with their approved deemed equity ratios.1015

1011 Exhibit 22570-X0783, paragraphs 43 and 45. 1012 Exhibit 22570-X0746, paragraphs 45-46, 49. 1013 Exhibit 22570-X0773, paragraph 12. 1014 Exhibit 22570-X0551, paragraph 185. 1015 Exhibit 22570-X0551, paragraph 203.



854. Mr. Thygesen also submitted that instead of showing mid-year debt balances, which are

the average of the opening and closing debt balances, the utilities should report the actual debt

balances at mid-year. He stated that while the mid-year debt balance is fairly easy to manipulate,

the actual debt balance at mid-year is more accurate.1016 Mr. Thygesen also suggested that

reporting actual debt/equity ratios are misleading when a utility deems debt levels in order to

reduce them to the approved percentage.1017

855. Mr. Buttke submitted that a requirement to report monthly debt and cash balances could

create a mechanism by which the utilities become incented to prioritize relatively small amounts

of visible savings over the far more valuable, but less visible, benefits of better risk

management.1018

856. AltaGas opposed Mr. Thygesen’s recommendation that the utilities report monthly cash

and cash equivalents, as well as monthly short-term debt balances, in Section 2 of their Rule 005

reports. AltaGas indicated that seasonality of its revenues and capital projects can affect

debt/equity ratios from month to month.1019 AltaGas and the ATCO Utilities submitted that any

proposed changes to Rule 005 are outside the scope of this GCOC proceeding.1020

857. ENMAX contended that reporting the monthly cash and cash equivalent balances would

create an incremental regulatory burden for all utilities, with no material offsetting benefit for

ratepayers. It submitted that requiring the monthly production of a cash report is unnecessary,

unduly burdensome, and is contrary to one of the fundamental PBR principles, which is to

reduce the regulatory burden.1021 ENMAX stated that as part of the quarterly review of its actual

capital structure, it must create a regulated income statement and balance sheet, and there are

costs associated with creating these financial documents. ENMAX argued that the cost of

producing these financial documents on a daily or monthly basis outweighs the benefits.

Commission findings

858. Mr. Thygesen’s recommendation that the utilities report monthly cash and cash

equivalent levels as well as monthly short-term debt balances as part of their Rule 005 reports

was to help ensure that the utilities are maintaining actual equity ratios in line with their

approved deemed equity ratios. In Section 11.1, the Commission denied Mr. Thygesen’s

recommendation that the affected utilities consistently attempt to align their actual equity ratios

with their deemed equity ratios by altering their cash management practices. On this basis, the

Commission finds there is no reason for the utilities to report monthly cash and cash equivalent

levels, or monthly short-term debt balances, as part of their Rule 005 reports.

859. However, the Commission agrees with Mr. Thygesen that the inclusion of deemed debt

levels as part of the information reported in Rule 005 can be misleading, and does not necessarily

portray an accurate calculation of the actual debt levels maintained by the utility. The

Commission therefore directs the utilities, as part of subsequent Rule 005 filings, to report actual

debt levels on their “Schedule of debt capital employed” and on their “Summary of mid-year

1016 Exhibit 22570-X0551, paragraph 182. 1017 Exhibit 22570-X0551, paragraphs 183-187. 1018 Exhibit 22570-X0749, A65. 1019 Exhibit 22570-X0783, paragraph 41. 1020 Exhibit 22570-X0783, paragraph 43. Exhibit 22570-X0746, paragraph 50. 1021 Exhibit 22570-X0773, paragraphs 8-9.



capital structure” schedule. In addition, the Commission directs the utilities to report the actual

cost rate from their “Schedule of debt capital employed” on their “Summary of return on rate

base schedule.”

11.3 Procedural issues

11.3.1 Use of aids to cross-examination

860. During the course of the oral hearing, a number of objections were raised with respect to

aids to cross-examination that counsel sought to put before various witnesses.1022

861. After ruling on a number of these objections, the panel chair expressed some concern on

behalf of the Commission with respect to the number of objections on proposed aids to cross-

examination:

More generally, in our correspondence leading up to this hearing and again in my

opening remarks, the Commission encouraged parties to make efficient use of time. The

Commission is concerned that the number of objections with respect to aids to cross is

becoming disruptive and not an efficient use of the Commission’s and parties’ time, the

cost of which is ultimately borne by ratepayers. We’re not looking to impede parties’

ability to explore the record or test the evidence, nor do we want to discourage counsel

from raising concerns about procedural fairness, but we have through our rulings

attempted to provide some guidance relative to the use of aids to cross.1023

…

We strongly encourage parties to revisit their planned use of aids to cross in light of the

guidance provided and otherwise work amongst themselves and have all the parties work

amongst themselves in an attempt to resolve some of these matters.1024

862. Repetitive objections and improper use of aids to cross-examination potentially

compromise regulatory efficiency and procedural fairness. Given the number and nature of the

objections raised during this proceeding, the Commission considers that a review of the proper

use of aids to cross-examination is warranted.

863. An aid to cross-examination should be used or referred to only if it assists in the

questioning of a witness on his or her evidence. An aid to cross-examination does not become

evidence in a proceeding, and cannot be relied on as proof of the matter that the aid to cross-

examination purports to prove. It is only what the witness says in relation to the aid to cross-

examination that becomes evidence. Therefore, an aid to cross-examination is generally of little

value to the Commission and will not be entered on the record if the witness being questioned

has neither contributed to the preparation of the document nor confirmed or adopted its content.

864. A valid aid to cross-examination must be relevant to the matter(s) before the Commission

and must be put to the witness(es) in a fair manner. While a document may be relevant, the party

or counsel who seeks to use the aid to cross-examination must also demonstrate the probative

1022 For example, Transcript, Volume 1, page 73; Transcript, Volume 1, page 74; Transcript, Volume 1, page 107;

Transcript, Volume 1, page 111; Transcript, Volume 1, page 126; Transcript, Volume 2, page 244; Transcript,

Volume 2, page 290; Transcript, Volume 4, page 727; Transcript, Volume 7, page 1436. 1023 Transcript, Volume 3, pages 480-481. 1024 Transcript, Volume 3, page 481.



nature of the document by tying it to the direct evidence or testimony of the witness(es). Fairness

involves sufficient time to review the document as well as allowing the witness to address

questions on it in the context of testing the witness’s evidence. The document’s connection to the

evidence and its intended use should be made clear.

865. Further, unless an aid to cross-examination is drawn directly from the witness’s direct

evidence or testimony, was prepared by that witness in another context, or provides updated or

supplementary information to the witness’s evidence, it is unfair and improper to ask the witness

to verify the information contained in the aid to cross-examination. To do otherwise would allow

the aid to cross-examination to be used as a means of introducing new evidence that could have

been put to the witness through written interrogatories or been included in a party’s filed

evidence.

866. It is also an inefficient use of the oral hearing time for counsel to be repeatedly objecting,

such as in instances where the relevance and fairness of a particular aid to cross-examination

does not appear to be truly in dispute. If a witness is unable to verify or comment on a particular

aid to cross-examination, the witness may so indicate. It is not always necessary for counsel to

interject, and counsel must be mindful to allow his or her witnesses to answer the questions fairly

put to them without interruption.

867. The Commission also requires that the relevant passages of longer aids to cross-

examination (five pages or more) be highlighted (Rule 001: Rules of Practice, Section 39.2).

This has been occasionally disregarded in past practice. Counsel should be aware that the

Commission may decide not to allow aids to cross-examination to be put to a witness that do not

comply with this requirement.

868. Finally, Section 39 of Rule 001 prescribes a minimum 24-hour notice period where a

party intends to use a document as an aid to cross-examination. However, parties are encouraged

to provide as much advance notice as is reasonably possible with a view to facilitating the early

identification, and possible resolution, of any issues in relation to the use of proposed aids to

cross-examination in advance of the hearing.

869. The Commission will consider if changes to its existing process regarding the use of aids

to cross-examination could address the above-noted concerns. The Commission may consider

these changes as an amendment to Rule 001, or it may direct parties to follow a revised process

in specific proceedings if the circumstances merit.

11.3.2 UCA concerns regarding process

870. In argument, the UCA submitted that there was a clear resource mismatch between the

utilities and interveners in this proceeding, which in its view was amplified by the procedure

adopted by the Commission, the extremely condensed nature of the process schedule and the

approach adopted by the utilities.1025 The UCA indicated that the main reason for reiterating

timing constraints in this proceeding is to hopefully ensure that sufficient time is allotted to both

counsel and witnesses in the next hearing process, so as to accommodate a thorough examination

of all information and to recognize that there is a clear mismatch in resources between

1025 Exhibit 22570-X0897.01, starting at paragraph 360.



interveners and the utilities.1026 The UCA recommended that the Commission ensure that the

process is commenced with sufficient time to allow a prospective determination of the cost of

capital while still allowing sufficient time to develop a full record, and that the Commission

reconsider a process whereby evidence and rebuttal evidence is filed simultaneously by the

utilities and interveners.1027

871. In reply, ENMAX objected to the concerns articulated by the UCA and submitted that the

sequential process used in this proceeding properly reflects the reality that it is the utilities that

bear the onus of proving that all aspects of their rates, including return and capital structure, are

just and reasonable.1028 In their joint reply, AltaLink, EPCOR and FortisAlberta submitted that

the UCA had advanced unfounded accusations that should be rejected, that the process for the

proceeding was fair and that any argument on future process is for the next GCOC

proceeding.1029 The ATCO Utilities and AltaGas indicated that while they support the UCA’s

first recommendation regarding establishing future GCOC timelines sufficient to develop a full

record and a prospective determination of cost of capital, they did not agree with the UCA’s

second recommendation.1030 The ATCO Utilities and AltaGas submitted that the cost of capital is

an important component of the utility revenue requirement, and that simultaneous filing of

evidence and rebuttal evidence would not be procedurally fair to the utilities.

872. The Commission acknowledges the concerns expressed by the UCA, but will not pre-

determine the process for a future GCOC proceeding in this decision. The UCA may highlight

any procedural concerns and requests with respect to process at the time of the next GCOC

proceeding for the Commission’s consideration at that time.

12 Implementation of GCOC decision findings

873. In the 2016 GCOC decision, the Commission approved an ROE of 8.5 per cent for all the

affected utilities, except ATCO Electric Transmission, on an interim basis for 2018 and any

subsequent year thereafter, unless otherwise directed by the Commission.1031 In the 2016 GCOC

decision, the Commission approved a deemed equity ratio of 37 per cent for AltaLink, ATCO

Electric Distribution, ATCO Gas, ATCO Pipelines, EPCOR, FortisAlberta, Lethbridge, Red

Deer and TransAlta; and a deemed equity ratio of 41 per cent for AltaGas, on an interim basis for

2018 and any subsequent year thereafter, unless otherwise directed by the Commission.1032 In

Decision 22121-D01-2016, the Commission approved an ROE of 8.5 per cent and a deemed

equity ratio of 37 per cent for ATCO Electric Transmission on an interim basis for 2018 and any

subsequent year thereafter, unless otherwise directed by the Commission.1033

874. In light of the Commission’s decision to maintain the existing approved ROE of 8.5 per

cent and deemed equity ratio of 37 per cent for cost-of-service utilities AltaLink, ATCO Electric

Transmission, ATCO Pipelines, EPCOR Transmission, Lethbridge, Red Deer and TransAlta, no

1026 Exhibit 22570-X0897.01, paragraph 371. 1027 Exhibit 22570-X0897.01, paragraphs 372 and 374. 1028 Exhibit 22570-X0909, paragraphs 74-76. 1029 Exhibit 22570-X0911, paragraphs 83-85. 1030 Exhibit 22570-X0918, paragraphs 276-278. 1031 Decision 20622-D01-2016, paragraph 628. 1032 Decision 20622-D01-2016, paragraph 628. 1033 Decision 22121-D01-2016, paragraph 10.



adjustment to any approved revenue requirements for 2018, 2019 and 2020 for these utilities will

be required with respect to ROE and deemed equity ratios as a result of this decision. As of the

date of this decision, ENMAX Transmission has no approved revenue requirement for 2018,

2019 or 2020. The Commission directs any utilities under cost-of-service regulation, being

AltaLink, ATCO Electric Transmission, ATCO Pipelines, ENMAX Transmission, EPCOR

Transmission, Lethbridge, Red Deer and TransAlta, who do not have Commission-approved

revenue requirements for any of 2018, 2019 and 2020, to incorporate the approved ROE and

deemed equity ratios as set out in this decision as part of their revenue requirement application(s)

for these years.

875. Any affected utility that has a Commission-approved revenue requirement under PBR for

2018 and subsequent years was required to use ROE and deemed equity ratio placeholders for

the purpose of the 2018 K-bar accounting test, until values were approved by the Commission on

a final basis. In light of the Commission’s decision to maintain the existing approved ROE of

8.5 per cent and deemed equity ratio of 37 per cent for all affected distribution utilities, other

than AltaGas, no adjustment to the revenue requirements to account for changes in approved

ROE or deemed equity ratios should be required for any of the affected utilities under PBR other

than AltaGas, as a result of this decision.

876. The Commission directs AltaGas to incorporate the approved deemed equity ratio of

39 per cent for 2018, 2019 and 2020 into all applicable rate proceedings and calculations that

rely on this approved deemed equity ratio, including the calculation of its base K-bar as part of

the next proceeding addressing adjustments to AltaGas’s 2017 notional rate calculations that

form the going-in rates for the 2018-2022 PBR term. To the extent that AltaGas, ATCO Gas,

ATCO Electric or FortisAlberta consider that this decision impacts the calculation of the income

tax expense included in their 2017 notional rate calculations that form the going-in rates for the

2018-2022 PBR term, this may similarly be addressed in the next proceeding considering any

required adjustments to their 2017 notional rate calculations.



13 Order

877. It is hereby ordered that:

(1) The final approved return on equity for AltaGas Utilities Inc., AltaLink

Management Ltd., ATCO Electric Ltd., ATCO Gas, ATCO Pipelines, ENMAX

Power Corporation, EPCOR Distribution & Transmission Inc., FortisAlberta Inc.,

the transmission operations of the City of Lethbridge, the transmission operations

of the City of Red Deer, and certain electricity transmission assets of TransAlta

Corporation, is set at 8.5 per cent for 2018, 2019 and 2020.

(2) The final approved deemed equity ratio for AltaLink Management Ltd., ATCO

Electric Ltd., ATCO Gas, ATCO Pipelines, ENMAX Power Corporation, EPCOR

Distribution & Transmission Inc., FortisAlberta Inc., the transmission operations

of the City of Lethbridge, the transmission operations of the City of Red Deer,

and certain electricity transmission assets of TransAlta Corporation, is set at

37 per cent for 2018, 2019 and 2020.

(3) The final approved deemed equity ratio for AltaGas Utilities Inc. is set at 39 per

cent for 2018, 2019 and 2020.



Dated on August 2, 2018.


(original signed by)

Mark Kolesar

Chair


Bill Lyttle

Acting Commission Member


Tracee Collins

Commission Member


Carolyn Hutniak

Commission Member



Appendix 1 – Proceeding participants

Name of organization (abbreviation) Company name of counsel or representative

AltaGas Utilities Inc. (AltaGas)

MLT Aikins LLP

AltaLink Management Ltd. (AltaLink)

Borden, Ladner Gervais LLP ATCO Electric Ltd. (ATCO Electric)

Bennett Jones LLP

ATCO Gas & Pipelines Ltd. (ATCO Gas) (ATCO Pipelines)

Bennett Jones LLP

Canadian Association of Petroleum Producers (CAPP)

Consumers’ Coalition of Alberta (CCA) Wachowich & Co.

Direct Energy Marketing Limited (Direct)

ENMAX Power Corporation (ENMAX) Torys LLP

EPCOR Distribution & Transmission Inc. (EPCOR)

Fasken Martineau Dumoulin LLP

EPCOR Energy Alberta GP Inc. (EEA)

FortisAlberta Inc. (FortisAlberta)

Industrial Power Consumers Association of Alberta (IPCAA)

Office of the Utilities Consumer Advocate (UCA)

Reynolds, Mirth, Richards & Farmer LLP The City of Calgary (Calgary)

McLennan Ross Barristers & Solicitors

TransAlta Corporation (TransAlta)



Alberta Utilities Commission Commission panel M. Kolesar, Chair B. Lyttle, Acting Commission Member T. Collins, Commission Member C. Hutniak, Commission Member Commission staff

K. Kellgren (Commission counsel) D. Reese (Commission counsel) D. Mitchell D. Ploof R. Lucas S. Crawford



Appendix 2 – Oral hearing – registered appearances

Name of organization (abbreviation) Name of counsel or representative

Witnesses

AltaGas Utilities Inc. and ATCO Utilities

R. Jeerakathil L. Smith, QC D. Sheehan

P. Carpenter B. Villadsen R. Buttke M. Stock G. Marghella

AltaGas Utilities Inc. (AltaGas)

R. Jeerakathil

AltaLink Management Ltd. (AltaLink), EPCOR Distribution & Transmission Inc. (EPCOR) and FortisAlberta Inc. (FortisAlberta)

R. Block, QC J. Liteplo J. Hulecki L. Ho J. Hennig L. Mason

R. Hevert D. Koch C. Lomore R. Drotar S. Chaudhary J. Sullivan A. Johnson

AltaLink Management Ltd. (AltaLink)

R. Block, QC

ATCO Utilities: ATCO Electric Ltd., ATCO Gas and Pipelines Ltd.

L. Smith, QC D. Sheehan

ENMAX Power Corporation (ENMAX)

D. Wood

J. Coyne A. Barrett J. McCoshen

EPCOR Distribution & Transmission Inc. (EPCOR)

J. Liteplo J. Hulecki

FortisAlberta Inc. (FortisAlberta)

L. Ho J. Hennig L. Mason

The City of Calgary (Calgary)

D. Evanchuk

H. Johnson

Consumers’ Coalition of Alberta (CCA)

J. Wachowich, QC

J. Thygesen D. Madsen

Office of the Utilities Consumer Advocate (UCA)

R. McCreary B. Schwanak

S. Cleary R. Bell



Alberta Utilities Commission Commission panel M. Kolesar, Chair B. Lyttle, Acting Commission Member T. Collins, Commission Member C. Hutniak, Commission Member Commission staff

K. Kellgren (Commission counsel) D. Reese (Commission counsel) D. Mitchell D. Ploof R. Lucas



Appendix 3 – Summary of Commission directions

This section is provided for the convenience of readers. In the event of any difference between

the directions in this section and those in the main body of the decision, the wording in the main

body of the decision shall prevail.

1. The Commission finds that because of the finite life of income tax loss carryforwards, as

opposed to the indefinite life of deductions such as capital cost allowance, the

conservative practice would be for utilities not to forecast income tax losses, but instead,

forecast the use of discretionary deductions such as capital cost allowance in order to

reduce forecast taxable income to zero. Accordingly, the Commission directs the utilities,

when forecasting income taxes, to only claim allowable deductions that will reduce the

taxable income to a maximum of zero. ............................................................ Paragraph 99

2. The Commission agrees with AltaGas and the ATCO Utilities that reporting the unfunded

FIT liability would have no bearing on their financial performance. However, given the

magnitude of the unfunded FIT balances that were forecast as of December 31, 2017, and

the Commission’s consideration that the calculation and reporting of this balance on an

annual basis would not require a significant amount of effort, the Commission directs the

ATCO Utilities, FortisAlberta, AltaGas and AltaLink to include their unfunded FIT

liability balance each year as part of their Rule 005 reports, beginning with the Rule 005

report for 2018, that will be submitted in 2019. The information provided should consist

of the unfunded FIT liability for the year being reported, as well as the previous year, and

the resulting difference. This information may assist the Commission in assessing the

level of potential credit metric relief that may be available if a utility were to apply to

adopt the FIT method. ................................................................................... Paragraph 102

3. The Commission notes that adjustments will be made to the distribution utilities’ going-in

PBR rates in future proceedings. For example, adjustments to going-in rates will be

required to reflect 2017 approved capital tracker amounts and to account for any

approved depreciation changes. The Commission directs AltaGas to revise the calculation

of its base K-bar to incorporate the findings in this decision as part of the next proceeding

addressing adjustments to AltaGas’s going-in PBR rates. To the extent that ATCO Gas,

ATCO Electric or FortisAlberta consider that this decision impacts the calculation of the

income tax expense included in 2018 going-in rates, this may similarly be addressed in

the next proceeding considering any required adjustments to their respective going-in

PBR rates. ..................................................................................................... Paragraph 135

4. However, the Commission agrees with Mr. Thygesen that the inclusion of deemed debt

levels as part of the information reported in Rule 005 can be misleading, and does not

necessarily portray an accurate calculation of the actual debt levels maintained by the

utility. The Commission therefore directs the utilities, as part of subsequent Rule 005

filings, to report actual debt levels on their “Schedule of debt capital employed” and on

their “Summary of mid-year capital structure” schedule. In addition, the Commission

directs the utilities to report the actual cost rate from their “Schedule of debt capital

employed” on their “Summary of return on rate base schedule.” .................. Paragraph 859

5. In light of the Commission’s decision to maintain the existing approved ROE of 8.5 per

cent and deemed equity ratio of 37 per cent for cost-of-service utilities AltaLink, ATCO

Electric Transmission, ATCO Pipelines, EPCOR Transmission, Lethbridge, Red Deer



and TransAlta, no adjustment to any approved revenue requirements for 2018, 2019 and

2020 for these utilities will be required with respect to ROE and deemed equity ratios as

a result of this decision. As of the date of this decision, ENMAX Transmission has no

approved revenue requirement for 2018, 2019 or 2020. The Commission directs any

utilities under cost-of-service regulation, being AltaLink, ATCO Electric Transmission,

ATCO Pipelines, ENMAX Transmission, EPCOR Transmission, Lethbridge, Red Deer

and TransAlta, who do not have Commission-approved revenue requirements for any of

2018, 2019 and 2020, to incorporate the approved ROE and deemed equity ratios as set

out in this decision as part of their revenue requirement application(s) for these years.

........................................................................................................................ Paragraph 874

6. The Commission directs AltaGas to incorporate the approved deemed equity ratio of

39 per cent for 2018, 2019 and 2020 into all applicable rate proceedings and calculations

that rely on this approved deemed equity ratio, including the calculation of its base K-bar

as part of the next proceeding addressing adjustments to AltaGas’s 2017 notional rate

calculations that form the going-in rates for the 2018-2022 PBR term. To the extent that

AltaGas, ATCO Gas, ATCO Electric or FortisAlberta consider that this decision impacts

the calculation of the income tax expense included in their 2017 notional rate calculations

that form the going-in rates for the 2018-2022 PBR term, this may similarly be addressed

in the next proceeding considering any required adjustments to their 2017 notional rate

calculations. .................................................................................................. Paragraph 876



Appendix 4 – Abbreviations

Abbreviation Name in full

2004 GCOC decision Decision 2004-052, Generic Cost of Capital

2009 GCOC decision Decision 2009-216, 2009 Generic Cost of Capital

2011 GCOC decision Decision 2011-474, 2011 Generic Cost of Capital

2013 GCOC decision Decision 2191-D01-2015, 2013 Generic Cost of Capital

2016 GCOC decision Decision 20622-D01-2016, 2016 Generic Cost of Capital

ACFA Alberta Capital Financing Authority

AESO Alberta Electric System Operator

AILP AltaLink Investments, L.P.

ALP AltaLink, L.P.

AltaGas AltaGas Utilities Inc.

AltaLink AltaLink Management Ltd.

ARCH autoregressive conditional heteroscedasticity

ATCO Electric ATCO Electric Ltd.

BHE Berkshire Hathaway Energy Company

Board Alberta Energy and Utilities Board

bps basis points

BYPRPM bond yield plus risk premium model

CAD Canadian dollar

CAD/USD Canadian dollar to the United States dollar

Calgary The City of Calgary

CAPM capital asset pricing model

CAPP Canadian Association of Petroleum Producers

CCA Consumers’ Coalition of Alberta

CIBC Canadian Imperial Bank of Commerce

CRA Canada Revenue Agency

CV coefficient of variation

CWIP construction work in progress

DACDA direct assigned capital deferral account

DBRS DBRS Limited

DCF discounted cash flow

EBIT earnings before interest and income taxes

EBITDA earnings before interest, income taxes, depreciation and

amortization

ECAPM empirical capital asset pricing model

ENMAX ENMAX Power Corporation

EPCOR EPCOR Distribution & Transmission Inc.

EPS earnings per share

EUI EPCOR Utilities Inc.

FFO funds from operations

FIT future income tax

FortisAlberta FortisAlberta Inc.

GARCH generalized form of ARCH

GCOC generic cost of capital

GDP gross domestic product



Abbreviation Name in full

GOC Government of Canada

GRA general rate application

GTA general tariff application

IBES Institutional Brokers’ Estimate System

I factor inflation factor

IMF International Monetary Fund

IR information request

LDC local distribution companies

Lethbridge City of Lethbridge

MC Alberta MidAmerican (Alberta) Canada Holdings Corporation

MERP market equity risk premium

Moody’s Moody’s Investor Services

MPR Monetary Policy Report

NAFTA North American Free Trade Agreement

NOI net operating income

O&M operating and maintenance

P/B price-to-book

PBR performance-based regulation

PP&E property, plant and equipment

PRPM predictive risk premium model

RBC Royal Bank of Canada

Red Deer City of Red Deer

ROE return on equity

S&P Standard & Poor’s

SML security market line

the affected utilities AltaGas Utilities Inc., the ATCO Utilities (ATCO Electric

Ltd., ATCO Gas and Pipelines Ltd.), ENMAX Power

Corporation, EPCOR Distribution and Transmission Inc.,

FortisAlberta Inc., City of Lethbridge, City of Red Deer

and TransAlta Corporation

the ATCO Utilities ATCO Electric Ltd., and ATCO Gas and Pipelines Ltd.

the Fed the Federal Reserve System

TransAlta TransAlta Corporation

TSX Toronto Stock Exchange

U.S. United States

UAD utility asset disposition

UAD decision Decision 2013-417, Utility Asset Disposition

UCA Office of the Utilities Consumer Advocate

USD U.S. dollar or United States dollar

VIX 30-day implied volatility of the S&P index (representing

the stock market in the U.S.)

VIXC 30-day implied volatility of the S&P/TSX 60 index

(representing the stock market in Canada)

WCS Western Canadian Select

WTI West Texas Intermediate

Decision 22570-D01-2018 2018 Generic Cost of Capital · 2018-08-03 · Decision 22570-D01-2018...

Documents

Transcript of Decision 22570-D01-2018 2018 Generic Cost of Capital · 2018-08-03 · Decision 22570-D01-2018...